# Nikhil Jathar — Full Content Export > Nikhil Jathar: serial entrepreneur, co-founder and CTO of AvanSaber. Acquired ventures, patent holder, Forbes Technology Council, building AI-native software. > Machine-readable full text of 36 published pages, generated at build time from https://www.nikhilj.com. Canonical source is the live site; each entry links its URL. --- ## Five rules for juggling multiple SaaS products · Nikhil Jathar URL: https://www.nikhilj.com/5-non-negotiable-productivity-rules-for-juggling-multiple-saas-platforms/ # Five rules for juggling multiple SaaS products I run eight products with a team of fifteen. Not with dashboards, with five simple rules that survive contact with a real, overloaded week. Nikhil Jathar15 November 20253 min read I run engineering across roughly eight products with a team of fifteen. People hear that and assume the answer is some elaborate system, a wall of dashboards and a color-coded life. The truth is closer to the opposite. The more I have to juggle, the fewer and simpler my rules have to be, because a complicated system is just one more thing to maintain. These are the five I actually keep, the ones that survived contact with real weeks. ## 1. One list, not eight Every product wants to be its own world, with its own backlog and its own urgency. The moment I let each one own a slice of my attention, I lost. So there is one list, across everything, ranked against everything else. InventoryPath does not compete with its own backlog. It competes with ERPClaw and with sleep. Forcing every priority onto a single ladder is uncomfortable, which is the point. It makes me admit what actually matters this week. ## 2. Context-switching is a tax, so batch by mode, not by product The instinct is to spend Monday on one product and Tuesday on the next. It feels organized and it is quietly ruinous, because the expensive switch is not between products. It is between modes of thinking. So I batch by mode. Deep building in one block regardless of which product it serves, shallow review in another, conversations in a third. Switching products inside a mode is cheap. Switching modes is the thing that empties the tank. ## 3. If it is not written down, it is not a decision With this many moving parts, my memory is a liability and a verbal agreement is a rumor. Anything that matters gets written, briefly, where the relevant person will see it. This sounds bureaucratic and is the opposite. Writing a decision down takes two minutes and saves the half-hour meeting where three people discover they remembered it differently. Across eight products, the scarce resource was never time. It was the ability to hold a clear thought without four others elbowing in. ## 4. Protect the maker's morning The single rule I break least: the first part of my day belongs to building, not to other people. I have written about the deep-work ritual at length, but the short version is that the work which compounds is fragile and slow to start, and if I let the morning fill with everyone else's priorities, the compounding work simply never happens. Afternoons are for the world. Mornings are mine. ## 5. Decide what you are willing to do badly This is the rule nobody likes. You cannot do eight things well at once, so the honest move is to choose, on purpose, what you will let slip this week, rather than failing at everything by accident. Some weeks the social channels go quiet. Some weeks a tidy refactor waits. Choosing your dropped ball in advance is the difference between a deliberate trade-off and a guilty surprise. I would rather decide what I am neglecting than discover it. That is the whole system. One list, batch by mode, write it down, defend the morning, choose your dropped ball. Five rules, because the sixth would be the first sign I had started managing my productivity instead of doing the work. When you are running everything you build, the goal is never to feel productive. It is to still be thinking clearly at the end of the week. NJ Nikhil Jathar “I run eight products and exactly one to-do list. The list wins.” --- ## Five reasons the 70s might be humanitys best decade · Nikhil Jathar URL: https://www.nikhilj.com/5-reasons-why-70s-was-the-best-period-for-humanity/ # Five reasons the 70s might be humanitys best decade A slightly ridiculous argument, made sincerely: the 1970s sat at a rare balance of modern tools and unclaimed attention that we have since optimized away. Nikhil Jathar24 December 20222 min read I want to make a slightly ridiculous argument with complete sincerity: the 1970s might have been the best decade humanity has had. I was not there for most of it, which possibly makes me the perfect person to romanticize it, free of the inconvenient memory of how it actually felt. But hear me out, because I think the case is better than nostalgia, and it says something about the strange moment we are in now. My argument is not that the 70s were comfortable or fair, because in many ways they were neither. It is that the decade sat at a rare and lucky intersection of forces that has not quite lined up the same way since. ## Why that particular decade The 70s had enough technology to make life genuinely better, and not yet enough to colonize every waking moment. People had television and cars and modern medicine, and they also had long uninterrupted hours that no glowing rectangle was competing for. The culture was extraordinary, music and film taking real creative risks, partly because the industry had not yet been fully optimized into focus-grouped sameness. And there was a sense of forward motion, of the future being something to build toward, that is harder to find now. It was, in other words, a moment with the benefits of modernity and not yet all of its costs. That balance is rarer than it sounds, and we tend not to notice when we are living through one until it is gone. ## What the nostalgia is really about I think when people my age romanticize the 70s, we are not really longing for that specific decade. We are longing for the balance it represents. Enough tools to be capable, not so many that we are owned by them. Enough connection to feel part of things, not so much that we never get a quiet minute. We spend a lot of energy now trying to recreate, deliberately and with great effort, a kind of attention that an ordinary 70s afternoon apparently just handed people for free. So no, I am not seriously suggesting we all go back. The 70s had plenty I am glad to live without. But I do think the decade got something right that we have since optimized away, and that the work of getting it back, carving out the uninterrupted hours, resisting the pull of the always-on, is worth doing on purpose. Sometimes the future is just remembering something good we did not realize we had. NJ Nikhil Jathar “We spend great effort now recreating an attention that an ordinary 70s afternoon handed people for free.” --- ## Accenture is becoming IBM, and so is consulting · Nikhil Jathar URL: https://www.nikhilj.com/accenture-is-becoming-ibm/ # Accenture is becoming IBM, and so is consulting Accenture beat earnings and hit a nine-year low the same day. I read it as the market re-rating the bodies-on-seats consulting model, not one quarter. Nikhil Jathar25 June 20266 min read I started my career at Accenture. Enterprise ERP, big teams, billing by the hour, the whole machine. So when Accenture hit a nine-year low this week, on an earnings beat, I did not read it as a bad quarter. I read it as the market quietly reclassifying what the company is, and by extension what the entire consulting model is worth. A fifteen percent drop on a beat is not noise. It is a re-rating. The market moved Accenture from a growth multiple toward a value multiple, and that swing is forty to fifty percent even when the fundamentals barely move. IBM has lived in that bucket for a decade: stable, profitable, and finished growing. The headline this week was "Accenture falls." The actual event was "Accenture becomes IBM." ## A re-rating, not a bad quarter Look at what actually broke, because it was not the present. Revenue was up six percent. EPS beat. Margins expanded. None of that is the story. The story is forward-looking: new bookings came in around 19.3 billion dollars against an expected 20.7, the next-quarter guide landed below consensus, and the full-year growth outlook got trimmed to three to four percent. The quarter was fine. The future got marked down, and markets pay for the future, so the stock paid for it. That is what a re-rating is. Not a verdict on last quarter, but a change in the category the market files you under. Growth companies earn a multiple on what they might become. Value companies earn a multiple on what they reliably are. The distance between those two numbers is enormous, and Accenture fell through it in a single session while its actual fundamentals barely moved. The label changed, not the business. ## It is bigger than one company If this were only Accenture, it would be a stock story. It is not. The same warning wiped roughly 1.35 lakh crore rupees off Indian IT before TCS or Infosys reported a single number, with Infosys touching a fresh 52-week low on the read-through alone. The market did not wait for their results. It had already decided that what it was repricing was a shape of company, not a single name: large teams, billed by the hour, sold by headcount. I know that shape from the inside. It is a pyramid, a few partners at the top and hundreds of juniors at the base, and the revenue underneath is, at its core, people times utilization times rate. For thirty years that was a wonderful business. You grew it by hiring. The more bodies you could bill, the more you earned, and the incentive and the model pointed the same way. That base is not going away tomorrow either. Most enterprises are slow not because they are foolish, but because scale, regulation, liability, and decades of accreted legacy make them slow. COBOL that still runs the back office, half-migrated SAP, ERP older than the people maintaining it: all of it pays for migrations and support for years. That inertia is real money, and it funds the incumbents for a long time. But "funded for years" is the IBM sentence exactly. It describes a company you hold for the dividend, not the growth. ## AI is a small-team technology, and their product is headcount Here is the structural problem. AI creates leverage with fewer, sharper people, not with armies billed by the hour. The whole promise is that a small team now does what a large one used to. That is the opposite of a pyramid. When your revenue is people times utilization times rate, and the technology's entire job is to eat the "people" term, the business you built and the moment you are standing in are pointed in different directions. This is why the CEO and the market cannot both be right. Julie Sweet says AI is a tailwind Accenture sells, and points to more than a hundred contracts worth over a hundred million dollars each. The tape says AI is a deflationary force that shrinks what the work is worth. One of those is a growth story and the other is a re-rating, and you cannot price both into the same stock. This week the market chose. ## The trust gap is the moat. It is also the clock. Now the part most hot takes get wrong, and the part I sit with longest, because it is close to the work I actually do. AI is a small-team technology, but enterprises will not let it be one inside their walls, and they are right to be careful. A regulated company does not hand a probabilistic system real access to real data on the strength of a good demo. The leap from a model that talks to one that acts, the shift ChatGPT plugins first made concrete, is exactly the leap a regulated enterprise is most cautious about. So they run small, contained experiments instead of going all in. And a small experiment still needs someone to scope it, wire it into twenty-year-old systems nobody dares replace, and own the risk when it touches production. That is consulting. That is SAP. For now, the trust gap is the incumbents' moat, and it is a real one. But look at what kind of moat it is. It produces pilots, not platforms. Maintenance, not rollouts. Caution billed by the hour. That is precisely the revenue mix the market pays a value multiple for, which means the moat and the downgrade are the same fact described twice. The trust gap does not save the model. It explains why the decline is slow, and it comes with a clock attached. The thing everyone assumes is protecting these firms, that enterprises do not trust AI with their data, is not a moat. It is a countdown clock. The clock runs out the day governance and guardrails make a regulated enterprise comfortable handing AI real access. When that happens, the small experiments do not stay small. They become large rollouts, run by small teams instead of armies. That bridge, the unglamorous scaffolding that makes AI safe to trust with real data and real authority, is the actual prize in enterprise AI. It is why we build and run the AI we recommend rather than only talk about it, and why, the two times we did file patents, we filed on the stubborn mechanism, not the demo. The bottleneck in enterprise AI was never capability. It is trust, and whoever makes trust cheap is the one who turns the pilot into the rollout. None of this is a next-quarter story. The legacy base and the trust gap fund the incumbents for years, and the bleed is slow by design. This is a medium-to-long re-rating, and the market did what it always does, which is price the ending early and then wait for everyone else to catch up. So the real question is not whether Accenture is dead. It plainly is not. The question is whether AI is a tailwind they sell or a deflationary force that shrinks what their work is worth. The CEO says the first. The tape voted for the second. I started on their side of this business and I now build on the other side of it, so I will say the quiet part out loud: I know which side I think the clock is on. That divide, between selling AI and building with it, is the one I keep circling back to in Fable and the Grid. NJ Nikhil Jathar “I started my career at Accenture. Now I help build the thing that puts a clock on the model. Make of that what you will.” --- ## AI on top of fifty-year-old systems · Nikhil Jathar URL: https://www.nikhilj.com/ai-on-top-of-fifty-year-old-systems/ # AI on top of fifty-year-old systems: what utilities taught me The real frontier of enterprise AI is layering assistants onto legacy utility systems without ripping them out. Why that unglamorous work matters most. Nikhil Jathar12 May 20263 min read The most interesting AI work I have seen in the last two years was not a chatbot or an image model. It was getting a useful answer out of a utility billing system that was designed before I was born. That is the unglamorous frontier, and it is where I think the real money and the real difficulty in enterprise AI actually live. I came up through enterprise ERP, so I have a specific bias here. When most people picture AI changing an industry, they picture greenfield: a clean new product, a clean new model, no legacy. But the industries that run the physical world, energy, water, the grid, do not get to start clean. They run on systems that have been accreting since the 1970s, and those systems are not going anywhere, because they work and the cost of being wrong is measured in people without power. ## The rip-and-replace fantasy The pitch a utility usually hears is that AI requires them to replace their core systems. This is mostly a fantasy, and an expensive one. You do not rip out the system of record for a million customers because someone demoed a clever assistant. The institutional knowledge baked into those platforms, the decades of edge cases and regulatory rules, is the actual asset. Throwing it away to chase a cleaner architecture is how transformation projects become cautionary tales. So the real work is the opposite of replacement. It is layering. You leave the system of record exactly where it is, and you put AI on top of it: assistants, analytics, and APIs that read from the old world and present a humane interface to the new one. Less heroic, far more useful. ## What that looks like in practice Take a utility call center. A customer calls because their bill tripled. Behind that simple question sits a tangle of meter reads, tariff rules, weather, and a fifty-year-old data model that was never designed to explain itself. The old way, an agent spends four minutes navigating screens. The useful AI way, a layer on top reads all of that and hands the agent, or the customer directly, a plain answer: here is why, here is what changed, here is what you can do. None of that requires replacing the billing engine. It requires understanding it well enough to translate it. Billing transparency, lower call volume, self-service that actually serves. The wins are concrete and boring, which is exactly why they are real. The hardest part of enterprise AI is not the model. It is having the patience to understand a system everyone else wants to throw away. ## Why this changed how I think about AI everywhere Working on this rewired my sense of which AI actually matters. The flashy demos assume a blank slate. The valuable work assumes the opposite: that the world is full of systems that are load-bearing, half-understood, and impossible to switch off, and that the job is to make them legible rather than to replace them. That is true far beyond utilities. Most of the economy runs on software somebody is afraid to touch. The teams that win with AI will not be the ones with the best models. They will be the ones willing to do the patient, unfashionable work of understanding the old system before they wrap it. We run our own stack for exactly this reason: it is the only way to learn that the boring part is the whole job. I find this genuinely exciting, which I realize is a strange thing to say about utility billing. But there is something honest about an AI problem where you cannot hide behind a nice demo, because a real person is waiting to find out why their bill went up. That constraint makes the work better. It usually does. NJ Nikhil Jathar “I get unreasonably excited about a correctly explained electricity bill. Someone has to.” --- ## Beyond the hype: three technologies I am excited about · Nikhil Jathar URL: https://www.nikhilj.com/beyond-the-hype-the-3-emerging-technologies-im-genuinely-excited-about/ # Beyond the hype: three technologies I am excited about The louder a technology is hyped, the more wrong I tend to be about it. Three emerging technologies that survived my own skepticism, and why they last. Nikhil Jathar29 July 20252 min read I have a fairly reliable personal indicator for technology. The louder a thing is hyped, the more wrong I usually turn out to be about it when I get excited, and the more right I am about the boring things nobody is shouting about. So when I say there are a few emerging technologies I am genuinely excited about, I mean it as a slightly nervous admission, because excitement is exactly the feeling that has misled me before. These three have survived my own skepticism, which is the only test I trust. ## 1. AI that makes old systems legible Almost all the attention goes to AI building new things from scratch. The work I find genuinely thrilling is the opposite: AI that sits on top of systems too old and too important to replace, and finally makes them explain themselves. I have written about this in the context of utilities, where a fifty-year-old billing engine can suddenly answer a human question. This is not glamorous and it is enormous, because the world runs on software nobody is willing to switch off. Making that software legible is worth more than another clever demo. ## 2. The collapsing cost of building The thing that has actually changed my daily work is how cheap it has become to build a first version of almost anything. I built a real server tool in a week that would have taken me a month not long ago. I am excited about this, and also wary, because cheap building is only a gift if you stay disciplined about what deserves to be built. The technology is not the interesting part anymore. The judgment about where to point it is. I have been wrong about nearly every overhyped technology and right about most of the boring ones. I have learned to trust the pattern. ## 3. Reliability as a first-class product The least exciting-sounding and most important shift is that as AI writes more software, the scarce thing becomes trust that the software actually works. This is why we built our own testing tooling and treat it as a real product, not a chore. When a model can produce a thousand lines in a minute, the bottleneck moves entirely to verification, and the teams that win will be the ones who made reliability a feature instead of an afterthought. Nobody hypes testing. Everybody needs it. ## What these three have in common None of them is the thing on the magazine cover. Each one is a layer of unglamorous plumbing that makes the flashy stuff actually usable: legibility for old systems, discipline for cheap building, trust for fast code. I have noticed that the technologies I stay excited about, years later, are almost always the plumbing. The cover stars tend to be gone by the next issue. So my filter has become simple, and a little unfashionable. I get suspicious of anything that is exciting mostly because it is new, and I pay attention to anything that is quietly making an old, real problem smaller. The hype points one way. The substance, in my experience, almost always sits somewhere quieter, doing the boring work that lasts. NJ Nikhil Jathar “I have been wrong about every overhyped technology and right about most of the boring ones.” --- ## DeepSeek R1: the underdog that shook up the AI world · Nikhil Jathar URL: https://www.nikhilj.com/deepseek-r1-the-underdog-thats-shaking-up-the-ai-world/ # DeepSeek R1: the underdog that shook up the AI world A model from a team nobody was watching broke the comfortable story that the AI race was already decided. Why the underdog mattered more than its scores. Nikhil Jathar27 January 20253 min read By the start of this year, a comfortable story had settled over the AI world. The race was effectively decided, the winners were the handful of giants with the most money and the biggest clusters, and the rest of us were just choosing which of them to build on. The earlier excitement, back when ChatGPT plugins first hinted at AI stepping out of the chat window, had cooled into a settled belief that scale alone decided everything. Then a model called DeepSeek R1 arrived from a team almost nobody outside the field had been watching, and the comfortable story fell apart in about a week. That is why it is the most interesting thing to happen in AI in a while, and not really because of the model itself. I am wary of hype, so let me be careful. R1 was not magic, and the breathless takes on both sides got it wrong. But it did something that mattered more than its benchmark scores. It proved the lead was never as safe as everyone had agreed to believe. ## Why an underdog model rattled everyone The shock was not only that R1 was good. It was that it was good while being radically cheaper to train and run, and open enough that anyone could pull it apart and learn from it. The entire premise of the established order was that frontier AI required resources only a few companies on earth could marshal. R1 was a loud, public counterexample, and counterexamples are how settled stories die. For a few days, the most powerful companies in the industry had to explain why their enormous moats were not, in fact, moats. That is a deeply healthy thing for any field to be forced to do. The most interesting thing in AI is almost never the biggest model. It is the one that was not supposed to be possible. ## What it actually means for builders For people like me, who build on this stuff every day rather than just talk about it, the lesson was practical and freeing. It meant the capability we depend on is commoditizing faster than the incumbents would like, that we are not permanently hostage to one vendor's pricing, and that betting your product on a single model being the only option was always more fragile than it looked. Optionality went up. That is good for everyone who builds and bad only for whoever was counting on a monopoly. It also fit a pattern I keep returning to. The thing worth paying attention to is rarely the loudest, best-funded contender. It is the quiet one doing more with less, which is usually where the genuinely interesting work hides. ## The bigger lesson, beyond one model R1 will be old news soon, replaced by whatever comes next, and that is exactly the point. The specific model matters less than what its arrival revealed: that in a field moving this fast, no lead is durable, no moat is permanent, and the assumption that the giants have already won is the most dangerous assumption a builder can hold. The race is not decided. It keeps not being decided, and an underdog from nowhere is always one good idea away from reminding everyone of that. I find that genuinely energizing, and a little destabilizing, which is probably the correct way to feel about any field worth working in. The moment we are certain we know who wins is the moment we have stopped paying attention to the person about to prove us wrong. It is the builder's posture I keep coming back to, the one I wrote about in Fable and the Grid: stay close to the work, stay loyal to no single tool. NJ Nikhil Jathar “The most interesting thing in AI is never the biggest model. It is the one that was not supposed to be possible.” --- ## ChatGPT plugins: AI steps out of the chat window · Nikhil Jathar URL: https://www.nikhilj.com/discover-the-power-of-chatgpt-plugins/ # ChatGPT plugins: AI steps out of the chat window ChatGPT plugins look like a small feature. They are a glimpse of where AI is heading: out of the chat window, into the world, doing rather than knowing. Nikhil Jathar12 May 20232 min read Something quietly important happened with AI recently, and it was not a bigger model. It was ChatGPT getting plugins, the ability to reach out of the chat window and actually do things in the world: browse, run code, talk to other services. On the surface it looks like a small feature. I think it is a glimpse of where this is all heading, and I want to write down why while it still feels new. Up to now, a chatbot has been a brilliant brain in a sealed jar. It could reason and write and explain, but it could not act. It only knew what it was trained on, and it could not check anything or change anything. Plugins start to remove the jar, and once you let a capable reasoning engine reach into real tools and live data, the nature of the thing changes. ## Why this matters more than a bigger model A smarter brain in a jar is still a brain in a jar. A slightly less smart brain that can use tools, look things up, and take actions is something else entirely, and far more useful for real work. We already know this from people. The most effective person is rarely the one who knows the most. It is the one who knows how to find what they need and get things done with it. Plugins point AI in that direction, toward doing rather than just knowing. That is the shift I find genuinely interesting. The frontier stops being how much the model knows and becomes what it can reliably do with the tools around it. ## The part to be careful about An AI that can only talk is safe in a boring, reassuring way. An AI that can take actions in real systems is more useful and more consequential, and both of those grow together. The same capability that lets it book the thing also lets it break the thing. I am genuinely excited about this direction, and I think the right response to it is care rather than either hype or fear. The interesting questions are about to stop being about intelligence and start being about trust, permissions, and what we are actually willing to let these systems do on our behalf. It is early, and plugins themselves may not be the form this takes. But the direction is the point. We are watching AI take its first real steps out of the chat window and into the world, and once a reasoning engine can act, very little stays the same. In the time since, that direction has only sharpened, from an underdog like DeepSeek R1 resetting who gets to compete, to what I learned putting these tools to work in a seven-day build of my own. NJ Nikhil Jathar “A smarter brain in a jar is still a brain in a jar.” --- ## Essays · Nikhil Jathar URL: https://www.nikhilj.com/essays/ # Essays 35 pieces on building, focus, leadership, and AI. Newest first. The 2025 onward work is where my current thinking lives. All 35 Building 6 Focus 4 Leadership 6 AI 9 Life 10 01 The four questions I got after the IRI talk Four questions came back after my IEEE IRI 2026 talk: why not just use ChatGPT, how invariants differ from unit tests, and whether it is really reuse. AI08 Aug 2026 · 6 min 02 Software Supply Chain Security in Plain Terms: SLSA, Sigstore, and in-toto SLSA, Sigstore, and in-toto without the jargon: how to prove a deployed artifact is what it claims to be, and was built the way you expect. Building02 Jul 2026 · 5 min 03 Accenture is becoming IBM, and so is consulting Accenture beat earnings and hit a nine-year low the same day. I read it as the market re-rating the bodies-on-seats consulting model, not one quarter. Leadership25 Jun 2026 · 6 min 04 Fable and the Grid: On Talking About AI, and Quietly Building With It Notes from a June where I spent 25 minutes telling US utilities how applied AI gets from pilot to production, then went home and built the next thing with it. AI16 Jun 2026 · 5 min 05 Eighteen years, then two in a row: RCB and the IPL RCB won the IPL again on May 31, back-to-back after eighteen years of winning nothing. On the encore, Kohli loyalty, and what the IPL became. Life02 Jun 2026 · 4 min 06 What we chose to patent, and why: inside ERPClaw and TailTest We filed patents on two products and not the rest. What was novel enough to protect, how a small studio thinks about IP, and what a patent really buys you. AI01 Jun 2026 · 3 min 07 AI on top of fifty-year-old systems: what utilities taught me The real frontier of enterprise AI is layering assistants onto legacy utility systems without ripping them out. Why that unglamorous work matters most. AI12 May 2026 · 3 min 08 How my co-founder and I actually split a company Ten years in, how Varun and I divide a company between a CEO and a CTO, where we overlap, and how we settle the arguments we cannot avoid. Leadership14 Apr 2026 · 3 min 09 What seven days of AI-assisted coding changed about how I build I built a server tool in a week with an AI pair. The build was the easy part. What it permanently changed about how I scope, delegate, and review. Building17 Mar 2026 · 4 min 10 Why we build and run the AI we recommend I am a founder who ships, not an advisor with slides. Why AvanSaber runs the AI it sells, and what operating our own software actually teaches us. Building18 Feb 2026 · 5 min 11 I built a Laravel Forge alternative in 7 days with AI I paid rent for two years on a tool I could build. So one December I gave myself seven days and an AI pair to make my own. Here is the honest version. Building04 Jan 2026 · 3 min 12 The deep-work ritual: how I structure my week I run engineering, strategy and leadership across several products. My only sane response is a deep-work ritual I defend with real stubbornness. Focus10 Dec 2025 · 3 min 13 Five rules for juggling multiple SaaS products I run eight products with a team of fifteen. Not with dashboards, with five simple rules that survive contact with a real, overloaded week. Focus15 Nov 2025 · 3 min 14 Work-life integration vs. balance: my philosophy Balance assumes two opposing weights and a virtuous person keeping them level. After a decade of running companies, I stopped believing in the scale. Leadership22 Oct 2025 · 3 min 15 Inbox zero is a myth. Inbox intentional is my reality. I chased inbox zero like it was a moral state, and produced almost nothing. Here is the messier, more useful habit I replaced it with. Focus12 Sept 2025 · 3 min 16 How I use productive downtime to prevent burnout The most productive thing I did last year was a week of near-nothing. Why rest is an input to the work, not a reward for it, and how I structure it. Focus22 Aug 2025 · 3 min 17 Beyond the hype: three technologies I am excited about The louder a technology is hyped, the more wrong I tend to be about it. Three emerging technologies that survived my own skepticism, and why they last. AI29 Jul 2025 · 2 min 18 The best leadership lesson I ever learned I learned it by getting it badly wrong during a scary week. Your team does not mirror what you say. It mirrors how you are, and managing that is the job. Leadership01 Jul 2025 · 3 min 19 My home-office tech stack: the gadgets I actually use People want a list of exciting gadgets. The honest one is short and boring: the few tools I use every day, and the expensive lessons left in the drawer. Building17 Jun 2025 · 2 min 20 The read-to-lead habit: five books that made me a better leader The worst leadership advice I ever read came from a leadership book. Why the books that actually made me a better leader were never about business at all. Leadership20 May 2025 · 2 min 21 Why I ask what is the why before any new project One question has killed more of my bad ideas than any plan or strategy session. Why I make every new project answer it clearly before we start building. Building30 Mar 2025 · 2 min 22 Mentorship is a two-way street: what young founders teach me I mentor younger founders and quietly get the better end of the deal. What they teach me that I could not learn any other way, and what I try to give back. Leadership13 Feb 2025 · 2 min 23 DeepSeek R1: the underdog that shook up the AI world A model from a team nobody was watching broke the comfortable story that the AI race was already decided. Why the underdog mattered more than its scores. AI27 Jan 2025 · 3 min 24 The future of work, part 2: human creativity in an AI world As AI gets better at the mechanical parts of work, it does not make human creativity less valuable. It quietly makes it the main thing worth paying for. AI05 Nov 2024 · 2 min 25 The future of work, part 1: navigating a new reality The future of work is not the vendor utopia or the headline apocalypse. It is a new reality to navigate carefully, by the people actually living it. AI24 Sept 2024 · 2 min 26 The funniest TV comedies since 2000 I switch off with comedy that is actually funny, not clever or important. My opinionated take on the best TV comedy since 2000, and why funny is so hard. Life16 Apr 2024 · 2 min 27 Why Its Always Sunny in Philadelphia is TVs best comedy An extreme position, defended sincerely: Its Always Sunny is the best comedy on television, and the cruelty that puts people off is exactly the point. Life16 Apr 2024 · 2 min 28 The social-media dopamine dilemma Social apps are engineered to hijack your attention with the same chemistry as a slot machine. Why willpower is the wrong tool, and what actually helps. Life11 Jan 2024 · 2 min 29 Five fun movies for your next movie night Some nights you do not want a film that changes you, you want one that is reliably fun. A small defense of rewatchable, low-stakes movie nights. Life07 Jan 2024 · 2 min 30 ChatGPT plugins: AI steps out of the chat window ChatGPT plugins look like a small feature. They are a glimpse of where AI is heading: out of the chat window, into the world, doing rather than knowing. AI12 May 2023 · 2 min 31 Rebellion on Mars: what frontier sci-fi taught me The Mars rebellion is one of sci-fis oldest stories. Underneath the spaceships, it is a builders story about earning the right to set your own terms. Life18 Apr 2023 · 2 min 32 Mens health in your 30s and 40s: skip the supplement hype Most of what is sold to men my age about health is noise. The three boring things that beat the supplement aisle, and where the pills actually fit. Life28 Dec 2022 · 2 min 33 Five reasons the 70s might be humanitys best decade A slightly ridiculous argument, made sincerely: the 1970s sat at a rare balance of modern tools and unclaimed attention that we have since optimized away. Life24 Dec 2022 · 2 min 34 Five underrated comedy TV shows worth finding The comedies that were brilliant and never got their due. Why great shows get overlooked, and the quiet joy of rescuing one for a friend. Life19 Dec 2022 · 2 min 35 In defense of Sachin Tendulkar The oldest thing I ever published online, from 2007: a young man defending Sachin Tendulkar against his critics, grammar fixed and argument intact. Life16 Nov 2007 · 3 min --- ## Fable and the Grid: On Talking About AI · Nikhil Jathar URL: https://www.nikhilj.com/fable-and-the-grid/ # Fable and the Grid: On Talking About AI, and Quietly Building With It Notes from a June where I spent 25 minutes telling US utilities how applied AI gets from pilot to production, then went home and built the next thing with it. Nikhil Jathar16 June 20265 min read Across June 8 and 9 I was in a room in Los Angeles, talking to a few hundred utility people about something I care about a lot and say out loud surprisingly rarely: how applied AI actually gets from a slide deck to a production system at a US utility. The session was called "Applied AI for Grid Modernization: From Pilot to Production at US Utilities," twenty five minutes, part of Smart Grids USA 2026. I will be honest about the part that made me grin backstage. The program had NERC, Schneider Electric, Siemens Energy, Dominion Energy, Lawrence Berkeley National Lab. Serious names who keep the lights on for tens of millions of people. And there I was, a founder who, the night before, had been arguing with an AI model about a database migration. The gap between "AI for the grid" as a panel topic and "AI as the thing I actually build with at 11pm" is the whole story, so let me tell it. ## The unglamorous version of applied AI Most talks about AI and infrastructure live at thirty thousand feet. Resilience. Optimization. Digital twins. All real. But the work I do is two feet off the ground. At AvanSaber we ship an unusually broad set of AI-native products, and a lot of that gets built shoulder to shoulder with Claude. Not a chatbot bolted onto an old app. The model in the loop while the thing gets made. So when I tell a utility audience that the leap is "pilot to production," I am not speaking in the abstract. The leap is the boring middle: the validation, the audit trail, the reversibility, the part where a probabilistic system has to live next to deterministic controls that absolutely cannot be wrong. That is the same discipline whether you are posting a journal entry in an ERP or switching load on a feeder. The demo is easy. The Tuesday after the demo is the job. ## Building with Fable Which brings me to the co-builder. The current model I lean on is Claude, in its Fable generation, and the honest experience of building with it is not the one in the keynote sizzle reels. It is quieter and better. It is less "AI writes your company for you" and more "a fast, tireless collaborator who never gets bored of the unglamorous part." A concrete example: the kind of project that used to be a quarter of engineering now fits in a week. I built a working Laravel Forge alternative in seven days with AI-assisted coding, the server provisioning, the deploys, the unglamorous plumbing, because the model never tired of the thousandth edge case and I stayed the one accountable for the result. That is the speed I was really describing on stage. The thing I keep noticing is that the experience rewards the same instinct I was on stage advocating: treat the AI as a capable junior who is brilliant and occasionally confidently wrong, put real guardrails around it, and keep a human accountable for the result. That is not a limitation. That is the operating manual. The teams who win with this are not the ones who trust it the most. They are the ones who verify the fastest. ## Why I bother saying it on a stage Utilities are exactly the kind of place where "move fast" is the wrong instinct and "move deliberately, with proof" is the right one. That is a feature, not a frustration. The most useful thing I think I said in those twenty five minutes was not a prediction. It was a permission slip: you do not have to choose between AI and control. The whole point of building the boring scaffolding is that you get to keep both. Then I flew home, opened the laptop, and got back to the boring scaffolding with my fast tireless collaborator. The divide between selling AI from a stage and building with it at 11pm is the one I keep coming back to, most directly in the Accenture piece. The grid and good software want the same thing in the end. Reliable, auditable, a little dull, and quietly doing their job while everyone argues about the future on a stage. NJ Nikhil Jathar “The future of the grid is boring on purpose. So is good software. I am fine with that.” --- ## How I use productive downtime to prevent burnout · Nikhil Jathar URL: https://www.nikhilj.com/how-i-use-productive-downtime-to-prevent-burnout/ # How I use productive downtime to prevent burnout The most productive thing I did last year was a week of near-nothing. Why rest is an input to the work, not a reward for it, and how I structure it. Nikhil Jathar22 August 20253 min read The most productive thing I did last year was a week where I did almost nothing. No big launch, no heroic sprint, just deliberate rest. I came back with a clearer head than three months of grinding had given me, and a decision I had been circling for ages made itself in about an hour. I tell this story to younger founders and watch them not believe me, because we have all been trained to think of downtime as the absence of work rather than a part of it. I used to believe that too. I burned out the way a lot of founders do, quietly and then all at once, and the lesson it left me with is the thing I want to write down here: rest is not the reward for the work. It is an input to it, and treating it as optional is how you slowly make yourself worse at your own job. ## Downtime is an input, not a luxury The work I am paid for is thinking, and thinking degrades when it never gets to stop. Tired judgment is bad judgment that feels exactly like normal judgment from the inside, which is what makes it dangerous. The decisions I made while running on empty were not just slower. They were worse in ways I could not see at the time, because the same exhaustion that produced them also dulled my ability to notice they were off. So I stopped framing rest as the thing I earn after the work and started treating it as part of the equipment that makes the work possible. You would not run a server at 100 percent utilization and call the headroom waste. The headroom is what keeps it from falling over. Tired judgment is bad judgment that feels completely normal from the inside. That is exactly what makes it dangerous. ## How I actually structure it The practical version is unglamorous. I protect real evenings and real weekends often enough that they are a pattern and not a rare event I have to be talked into. I take proper breaks between intense stretches rather than rolling one straight into the next, because the recovery is where the previous push actually consolidates into something I can use. And I treat genuinely unstructured time, a walk with no podcast, an afternoon with no plan, as productive rather than indulgent, because that is reliably where my least obvious ideas show up. This sits right next to how I think about integrating work and life and why I am so strict about defending my deep-work hours. The deep hours and the empty hours are not opposites. They are the same system. One does not function without the other, and I learned that the expensive way. ## The discipline of stopping The hard part is that stopping requires more discipline than continuing. Continuing is the default, especially when you run things you care about and there is always one more reasonable thing to do. Choosing to stop while there is still work undone, on purpose, trusting that you will be better for it, goes against every instinct that got you here. But the instinct is wrong, or at least incomplete. The founders I watch last are not the ones who can grind the longest. They are the ones who learned, usually the hard way, how to stop before the grinding starts costing them the very judgment they are grinding to protect. I treat my rest like a deductible expense now. It is not the fun part of the budget, and skipping it always looks affordable in the moment. The bill for skipping it, when it finally arrives, never is. NJ Nikhil Jathar “I treat rest like a deductible expense. Skipping it always looks affordable until the bill arrives.” --- ## How my co-founder and I actually split a company · Nikhil Jathar URL: https://www.nikhilj.com/how-my-co-founder-and-i-split-a-company/ # How my co-founder and I actually split a company Ten years in, how Varun and I divide a company between a CEO and a CTO, where we overlap, and how we settle the arguments we cannot avoid. Nikhil Jathar14 April 20263 min read For the first few years, AvanSaber had no org chart, because an org chart of two people is just two names and a line, and we already knew the line. Varun ran the company. I built the products. We are now past a decade of that arrangement, and the line has not really moved, which I think is the most useful thing I can tell another founder about partnership. The split that works is the one you can still describe in a single sentence ten years later. People romanticize co-founders, or they tell horror stories. The truth in our case is quieter than both. We are not best friends who finish each other's sentences, and we have never thrown a chair. We are two people with genuinely different wiring who decided, early and on purpose, to stop competing for the same job. ## The clean split: business and building Varun is the CEO. He owns the company as a business: the customers, the commercials, the direction, the parts that involve persuading the outside world that any of this is worth paying for. I am the CTO. I own the company as a machine: what we build, how it is built, and whether it stays standing when someone leans on it. When a customer asks a question, the answer almost always belongs cleanly to one of us, and we both know which. This sounds obvious. It is not how most early companies actually behave. Two technical founders will quietly fight over architecture. Two commercial founders will quietly fight over the deck. The thing that saved us was that neither of us wanted the other's job. I have no desire to run a sales call. He has no desire to review a database schema. That lack of overlap is not a weakness in the partnership. It is the partnership. ## Where the split gets messy It is clean until it is not, and the place it stops being clean is always the same: anything that is both a business decision and a building decision at once. Which product gets the next two engineers. Whether we take on a customer whose needs will bend the roadmap. How much of a quarter we spend on something that pays off in two years or not at all. Those calls live exactly on the seam between us, and early on we were bad at them. I would treat a roadmap question as purely technical. He would treat it as purely commercial. We were both half right, which is the most annoying way to be wrong. The best decisions we make are the ones neither of us could have reached alone, and was certain about until the other pushed back. ## How we actually settle it What we landed on is unglamorous. For anything on the seam, the person whose domain carries more of the risk gets the final call, and the other one gets a real, listened-to objection first. If a decision will mostly be felt by customers, it is his. If it will mostly be felt at 2am by whoever is on call, it is mine. And whoever does not get the call has agreed, in advance, to disagree and commit rather than relitigate it for a month. The trust that makes this work is not vague good feeling. It is specific. I trust that he is not optimizing for looking right. He trusts that I am not gold-plating something for my own satisfaction. We earned that read on each other slowly, by being wrong in front of each other and not making it weird. ## What a decade taught me The thing I did not expect is how much a good co-founder functions as a second kind of judgment, not just a second pair of hands. A lot of what I have learned about backing people and learning from them started with Varun, because he was the first person whose disagreement I learned to treat as information rather than friction. If you are picking a co-founder, I would worry less about shared passion and more about a clean division of who owns what, plus enough trust to lose an argument gracefully. The shared passion fades on the hard Tuesdays. The division of labor is what gets you to Wednesday. We run everything we build, which means there is always plenty to disagree about. Ten years in, the disagreements are still the most productive part of my week. NJ Nikhil Jathar “Varun runs the company. I run the parts that page me at 2am. We are both convinced we got the better deal.” --- ## I built a Laravel Forge alternative in 7 days with AI · Nikhil Jathar URL: https://www.nikhilj.com/i-built-a-laravel-forge-alternative-in-7-days-using-ai-assisted-coding/ # I built a Laravel Forge alternative in 7 days with AI I paid rent for two years on a tool I could build. So one December I gave myself seven days and an AI pair to make my own. Here is the honest version. Nikhil Jathar4 January 20263 min read For about two years I paid a monthly fee for a tool that deployed my servers. It was a good tool. I had no real complaint, except the quiet one that never goes away for people like me: I knew, more or less, how it worked, and I resented paying rent on a thing I could probably build. So one December, between the holidays, I decided to find out if "probably" was doing too much work in that sentence. I gave myself seven days. The result is SiteKit, a server and deployment tool we now run and have open-sourced. I want to tell the honest version of how it came together, because the headline, built in a week with AI, is true and also slightly misleading in a way that matters. ## The bet, and why a week I did not pick seven days because it was realistic. I picked it because a deadline that is slightly unreasonable is the only kind that prevents a side project from quietly becoming a six-month project. The goal was not a perfect product. It was a thing that worked, that I understood completely, and that I could put on a real server without wincing. I had an AI model open the entire time, as a pair. Not as a magic wand, as a fast and tireless junior who had read everything and remembered none of the consequences. By then AI had already grown past the chat box, from ChatGPT plugins reaching into real tools to models that could draft and run real code, so the question was never whether it could help but how much of the judgment stayed mine. ## The middle three days felt like cheating The first day was slow, because the first day is always thinking, not typing. Once I knew the shape of the thing, the middle stretch was genuinely startling. The boilerplate, the API wiring, the second and third versions of a function I was still figuring out, all of it came in minutes instead of hours. If you measured those days by output, it looked like I had hired a team. But output is a liar's metric, and I knew it even as it felt great. Every line the model wrote was a line I would later have to understand under pressure, because a server tool that you do not fully understand is not a tool. It is a future outage with your name on it. The model gave me a week's worth of code in three days, and then quietly handed me the bill: now understand all of it. ## The hard part was exactly where you would expect The trouble started when the tool met a real server. Permissions, edge cases, the difference between a thing that works on my machine and a thing that works at two in the morning when something has gone wrong and a real site is down. The model was confidently useless here, because these problems are not in the textbook. They are in the scar tissue, and scar tissue does not train well. So days five and six were old-fashioned. Reading my own code as if a stranger had written it, which in a sense one had. Deleting the clever parts. Adding the boring safety I would want at 2am. This is the work that does not demo well and decides everything. ## What shipped, and what it cost On the seventh day it worked. Not gloriously, but really. It deployed, it monitored, it did not fall over when I was unkind to it. We have been running it ever since, and putting it out in the open was its own small lesson in humility, because open source means strangers can read the parts you rushed. I have written separately about what those seven days permanently changed in how I build everything now. The short version is that AI moved the work rather than removing it. It made the typing cheap and made judgment the entire job. I would do it again in a heartbeat, which is why I am now strict about asking why before I start, because the new danger is not that building is hard. It is that building just got easy enough to do thoughtlessly. And the capability underneath keeps getting cheaper, as an underdog like DeepSeek made plain, which only sharpens the point: the scarce thing was never the code, it was knowing what to build and why. The thing I am proudest of is not the speed. It is that I can still explain every meaningful line, a year later, because in the only way that counts, I did write it. The model just typed faster than I can. NJ Nikhil Jathar “Seven days to build it. Seven months to stop mentioning that it took seven days.” --- ## In defense of Sachin Tendulkar · Nikhil Jathar URL: https://www.nikhilj.com/in-defense-of-sachin-tendulkar/ # In defense of Sachin Tendulkar The oldest thing I ever published online, from 2007: a young man defending Sachin Tendulkar against his critics, grammar fixed and argument intact. Nikhil Jathar16 November 20073 min read This is the oldest thing I have ever published online. I wrote it in 2007, on a Blogspot page, defending Sachin Tendulkar against the people who had decided he was selfish, or finished, or somehow not a match-winner. I have fixed the grammar and removed the dead links, but I have left the argument almost exactly as the younger version of me made it, because I still mostly agree with him. The case against Sachin, repeated endlessly back then, came in three parts: that India lost whenever he scored a century, that he cared only about his records, and that he was not a team player. All three are wrong, and the reason is simple. Cricket is a team game, and you cannot blame one man for the failure of ten others. ## The Mohali example Take the second one-day international against Pakistan at Mohali. Tendulkar scored a composed 99 and was out in the 26th over with India at 179 for 2, the run rate close to seven an over. From there, a side with any depth finishes near 340. India managed 321 for 9, because almost everyone after him failed, and then lost the game by a narrow margin. That scorecard was offered, that week, as evidence against Sachin. It was actually evidence against everyone else. You cannot lose a fifty-over match on one man's innings. He had done his job and walked off with the platform built. What the others did with it was not his mistake, and calling it one only revealed who was keeping score in bad faith. A soldier is not judged by the weapon he carries or the battles he wins, but by the character with which he fights. ## More than a match-winner People with genuinely dubious credentials would still ask: is Sachin even a match-winner? He is far more than that. For more than a decade he was the hope that held a fragile team together, walking out with a purpose the rest of the order rarely matched. Do not judge him only by the runs. He raised a whole generation of young cricketers who would carry Indian cricket for the next twenty years, and that is a larger kind of winning than any single chase. The real problem was never his ability. It was our expectation. Every single time he walked out to bat, an entire country expected a hundred. He is, in the end, a human being who has conquered the game more often than anyone had a right to ask. Judging him against perfection and then calling the gap a flaw says far more about us than about him. ## Old is gold If you want one innings, take Sharjah in 1996. Tendulkar's 118 carried India past 300 for the first time in one-day cricket, in a record second-wicket stand with Navjot Sidhu, after he had scored just three runs in two earlier defeats. He played fast bowling and spin with the same ease, then took two wickets with his odd mix of flippers and top-spinners, and India won by 28 runs. That was the whole player in one afternoon. I will give the last word to the one man who outranks every critic on this subject. Don Bradman watched Tendulkar bat on television and called his wife into the room, because the technique, the compactness, the stroke production, all of it reminded him of how he himself used to play. When Bradman sees himself in you, the argument is over. It was over in 2007, and it is more over now. NJ Nikhil Jathar “My first piece of internet writing, from 2007. I fixed the grammar and kept the conviction.” --- ## Inbox zero is a myth. Inbox intentional is my reality. · Nikhil Jathar URL: https://www.nikhilj.com/inbox-zero-is-a-myth-inbox-intentional-is-my-reality/ # Inbox zero is a myth. Inbox intentional is my reality. I chased inbox zero like it was a moral state, and produced almost nothing. Here is the messier, more useful habit I replaced it with. Nikhil Jathar12 September 20253 min read For a while I chased inbox zero like it was a moral state. Empty inbox, clear conscience. I would spend the back end of a day filing, archiving, and replying just to watch the number tick down to nothing, and I would feel virtuous, and I would have produced almost nothing of value. It took me too long to notice the obvious thing: a clean inbox and a good day are not the same achievement, and chasing the first one was quietly costing me the second. So I gave up on inbox zero, on purpose, and replaced it with something I find far more useful. I call it inbox intentional, and the difference is not cosmetic. It changed how I spend the most easily wasted hours of the week. ## Why zero is the wrong target Inbox zero optimizes for the wrong number. It treats every message as a task to be cleared, which means the people who email you the most get to set your agenda, and the urgent quietly defeats the important every single time. An empty inbox tells you that you have responded to the world. It tells you nothing about whether you did anything that mattered. The deeper trap is that clearing the inbox feels like work, because it is effortful and it produces a visible result. But effort plus a visible result is exactly how busywork disguises itself as progress. I was getting a small hit of accomplishment for doing the least valuable thing on my plate. An empty inbox proves you answered the world. It says nothing about whether you did anything worth answering for. ## What intentional looks like instead Inbox intentional starts from a different question. Not "how do I clear this," but "what here actually deserves my attention, and when." A few rules carry most of it. I process email in two or three fixed windows a day, never continuously, because an inbox left open is just a tap dripping interruptions onto whatever I am really trying to do. Most messages get a quick decision rather than a careful reply, because most messages do not warrant a careful reply and pretending otherwise is how the day disappears. And the small set of emails that genuinely matter get real thought, often later, on purpose, rather than a fast answer squeezed between two other things. The inbox stops being a scoreboard and goes back to being what it always should have been: a queue of inputs that I triage on my schedule, not a list of obligations that triage me. ## The point was never tidiness My inbox is rarely at zero now, and I have made a deliberate peace with that. There are usually a few dozen things sitting in it that I have consciously decided are not worth a response yet, or ever. That is not a backlog. That is a decision, made once, instead of a guilt I re-litigate every time I glance at my phone. This is really the same principle that runs through how I protect my mornings and how I juggle several products at once. The scarce resource is attention, and attention spent reacting is attention not spent thinking. Inbox zero is a tidy way to spend a day reacting. Inbox intentional is a slightly messier way to spend it on the things that actually move. The empty inbox was never the achievement. It just looked like one, which is exactly what made it so easy to keep chasing. NJ Nikhil Jathar “Inbox zero is a tidy way to spend a whole day achieving nothing.” --- ## Mens health, 30s and 40s: supplement hype · Nikhil Jathar URL: https://www.nikhilj.com/maximizing-mens-health-in-their-30s-and-40s-with-vitamin-supplements/ # Mens health in your 30s and 40s: skip the supplement hype Most of what is sold to men my age about health is noise. The three boring things that beat the supplement aisle, and where the pills actually fit. Nikhil Jathar28 December 20222 min read I am in the stretch of life, somewhere in the middle, where your body stops giving you things for free and starts sending invoices for how you treated it in your twenties. As a founder who has spent years prioritizing work over almost everything, including sleep, I have had to actually think about health rather than assume it. And the first thing I learned is that most of what is sold to men my age about it is noise. The supplement industry in particular has perfected the art of selling certainty about uncertain things. Walk into any shop and you will be told a dozen pills will fix your energy, your focus, your everything. The honest truth, which sells nothing, is that for most reasonably healthy people the expensive cabinet of supplements does far less than three boring things that cost almost nothing. ## The boring things that actually work Sleep, real sleep, is the closest thing to a miracle drug I have found, and the one I spent years foolishly treating as optional. Movement, not heroic exercise, just regular honest movement, does more for my energy and my head than any pill ever has. And eating like an adult most of the time, without turning it into a religion, handles most of the rest. None of this is exciting. All of it works, and it keeps working long after the supplement of the month is forgotten. This is the same lesson I keep relearning in every part of life. The unglamorous fundamentals, done consistently, beat the exciting shortcut almost every time. ## Where supplements actually fit I am not against supplements entirely. If a blood test shows you are genuinely low in something, fixing that is just sensible, and a couple of well-evidenced basics are reasonable insurance. The trap is the other way around: taking a fistful of pills to compensate for sleep you are not getting and movement you are not doing. That is paying for a patch instead of fixing the leak, and the body is not fooled. I think about this the way I think about protecting my recovery generally. Health in your thirties and forties is not bought in a supplement aisle. It is built, slowly and unglamorously, out of sleep and movement and not lying to yourself, and the sooner you stop looking for the shortcut, the sooner you can start on the boring things that actually pay. NJ Nikhil Jathar “Health in your forties is not bought in a supplement aisle. It is built out of sleep, movement, and not lying to yourself.” --- ## Mentorship: what young founders teach me · Nikhil Jathar URL: https://www.nikhilj.com/mentorship-is-a-two-way-street-what-i-learn-from-young-founders/ # Mentorship is a two-way street: what young founders teach me I mentor younger founders and quietly get the better end of the deal. What they teach me that I could not learn any other way, and what I try to give back. Nikhil Jathar13 February 20252 min read I spend a fair amount of time mentoring younger founders, and I should admit something that probably undermines the whole noble premise: I am almost certainly getting the better end of the deal. I started doing it to give something back, which is the respectable reason everyone offers. I have kept doing it because I learn more in those conversations than the people I am supposedly teaching, and at some point pretending otherwise felt dishonest. The phrase "mentorship is a two-way street" gets used so often it has gone soft, so let me be specific about what I actually mean. The founders I advise teach me things I genuinely could not learn any other way, because they are standing somewhere I cannot stand anymore. ## They can see what I have stopped seeing After enough years, you accumulate a kind of expensive blindness. You know how things are done, and that knowledge quietly closes doors you no longer notice are there. A founder two years into their journey has not learned yet what is supposedly impossible, so they routinely ask the question I have been trained not to ask, and often it is the right question. Half the time my honest internal reaction is, I have no idea why we don't do it that way, and I have to go find out. They are also closer to the current ground truth than I am. They are using tools I have not had to learn, hiring in a market I last saw years ago, and reading the present rather than remembering the past. I get a free, continuous education in how the world actually works right now, in exchange for opinions I would happily give for nothing. I started mentoring to give back. I kept doing it because I was quietly learning more than I was giving, and it felt rude to stop. ## What I try to actually give In return, the useful thing I can offer is rarely answers. It is pattern recognition and perspective on time. I have made many of the mistakes they are about to make, so I can sometimes say, gently, that the thing keeping them awake will not matter in a year, and that the thing they are ignoring will. I cannot make their decisions, and I try hard not to. I can occasionally widen the lens. Some of what I have learned about staying composed is most useful precisely when handed to someone in the middle of their first real crisis. ## Why I keep the door open This is not so different from how my co-founder and I keep each other honest. The value comes from the friction of a different vantage point, from someone who is not trapped inside your assumptions. Young founders are not a charity case I tolerate out of duty. They are one of the best ongoing sources of fresh thinking I have, and the day I start believing I am only the one teaching is the day I should probably stop, because I will have stopped paying attention to the better half of the conversation. NJ Nikhil Jathar “I started mentoring to give back. I kept doing it because I was learning more than I gave.” --- ## My home-office tech stack: the gadgets I actually use · Nikhil Jathar URL: https://www.nikhilj.com/my-home-office-tech-stack-the-gadgets-i-actually-use/ # My home-office tech stack: the gadgets I actually use People want a list of exciting gadgets. The honest one is short and boring: the few tools I use every day, and the expensive lessons left in the drawer. Nikhil Jathar17 June 20252 min read People ask about my home office setup fairly often, usually hoping for a list of exciting gadgets. I always feel slightly guilty disappointing them, because the honest list is shorter and more boring than anyone wants it to be. After years of buying the thing that promised to make me more productive, I have ended up with a small set of tools I actually use every day, and a drawer of expensive lessons about the ones I do not. The pattern I eventually noticed is that the gadgets that survived in my routine all have one quality in common. I have stopped noticing them. The ones I got rid of are the ones that kept asking for my attention. ## The short list that earned its place A genuinely good keyboard and a genuinely good chair, because I spend most of my life touching one and sitting in the other, and being cheap about your primary interfaces is a false economy I learned the slow way. A single large monitor, not the wall of three I once thought I needed, because more screen turned out to mostly mean more places for my attention to scatter. Good headphones, less for the audio than for the signal they send the rest of the house that I am, for the next hour, unavailable. That is close to the whole list. None of it is exciting. All of it disappears in use, which is exactly why it stayed. The best gadget I own is the one I have completely stopped noticing. The worst were the ones that never stopped asking to be noticed. ## What I stopped using, and why The casualties are more instructive than the survivors. The smart gadgets that needed their own app. The note-taking system so elaborate that maintaining it became a second job. The standing desk I was sure would change my life and now use, honestly, as a shelf. Every one of them failed the same test. It added a small, constant tax on my attention in exchange for a benefit I could rarely measure. I think about tools now the way I think about the software we build. A good one removes friction and then gets out of the way. A bad one adds a feature and a notification and a reason to think about it. With a setup, as with a product, the highest praise is that you forget it is there. ## The actual principle If there is a rule underneath all of this, it is that a tool should serve the work, not become the work. The point of my setup is to make it as easy as possible to do the deep work that matters and as hard as possible for anything to interrupt it. Almost everything I kept does one of those two things. Almost everything I dropped quietly did the opposite while pretending to help. So when someone asks for my list of must-have gadgets, the truthful answer is anticlimactic. Buy fewer things, buy good versions of the few you actually touch all day, and be ruthless about anything that wants your attention more than it earns it. The most productive setup I have ever had is also the most forgettable, and that is not a coincidence. NJ Nikhil Jathar “The best gadget I own is the one I have completely stopped noticing.” --- ## Five fun movies for your next movie night · Nikhil Jathar URL: https://www.nikhilj.com/my-top-5-fun-movies-to-watch-on-your-next-movie-night/ # Five fun movies for your next movie night Some nights you do not want a film that changes you, you want one that is reliably fun. A small defense of rewatchable, low-stakes movie nights. Nikhil Jathar7 January 20242 min read Some nights you do not want a film that changes you. You want a film that is simply, reliably fun, the cinematic equivalent of a good meal you have had before and know you will enjoy. I have a small mental list of these, the movies I put on when the week has been long and my brain is full and I want to be entertained without being asked for anything in return. Here is roughly what is on it. My only criterion is rewatchability. A film can be a masterpiece and exhaust me. The movies on this list are the opposite. They are the ones I can drop into at any point, that ask nothing, and that I never regret choosing. That is a real and underrated quality, and the people who make these films well rarely get the credit the prestige directors do. ## What makes a great fun movie The best fun movies share a few traits. They have momentum, so you are never bored. They have charm, usually carried by actors who are clearly enjoying themselves, which is infectious. And they do not take themselves too seriously, which paradoxically takes enormous skill, because the difference between effortless fun and trying-too-hard is razor thin and the audience feels it instantly. The heist films, the smart action comedies, the adventure films with a sense of humor about themselves. These are the genres that reward a tired brain. They are built to be enjoyed, and being genuinely enjoyable is harder to engineer than being impressive. ## A small defense of low-stakes pleasure There is a quiet snobbery that treats fun as a lesser ambition than meaning, and I have never bought it. Making something a million people genuinely enjoy, again and again, is a serious achievement. As someone who builds things people use every day, I have a lot of respect for work whose only goal is to make someone's evening better, and that succeeds at it. That is not a small thing. It is most of what any of us are actually trying to do. So the next movie night, give yourself permission to skip the film you think you should watch and put on the one you know you will enjoy. Fun is not a guilty pleasure. On a long Friday, it might be the most sensible choice you make all week. NJ Nikhil Jathar “Fun is not a guilty pleasure. On a long Friday it might be the most sensible choice you make.” --- ## Eighteen years, then two in a row: RCB and the IPL · Nikhil Jathar URL: https://www.nikhilj.com/rcb-back-to-back-ipl-2026/ # Eighteen years, then two in a row: RCB and the IPL RCB won the IPL again on May 31, back-to-back after eighteen years of winning nothing. On the encore, Kohli loyalty, and what the IPL became. Nikhil Jathar2 June 20264 min read On the thirty-first of May, Royal Challengers Bengaluru won the IPL final in Ahmedabad, chased down Gujarat Titans with twelve balls to spare, and Virat Kohli walked off unbeaten on 75. It was their second title in a row. I want to sit with that small phrase, second in a row, because a year ago it would have sounded like science fiction. For eighteen years, RCB had won nothing at all. I am not a neutral here and will not pretend to be. But you do not have to love the team to find the story useful, because it is one of the cleanest examples I know of something I think about constantly as a builder: the gap between finally succeeding and proving the success was not an accident. ## Eighteen years of being the joke For most of the IPL's life, RCB were the league's lovable disaster. They had the biggest names, the loudest fans, and a real genius for losing in the most painful way available on the night. They reached the final in 2009, in 2011, and in 2016, and lost all three. Their own supporters turned their hope into a punchline, chanting "Ee Sala Cup Namde," this year the cup is ours, every single season while knowing it almost certainly was not. Eighteen years is not a slump. It is a personality. Then in 2025, after more than six thousand days, they finally won, beating Punjab Kings by six runs. The release was extraordinary, the specific kind of joy that only that much stored disappointment can produce. If the story had ended there it would still have been a good one. A long wait, a deserved reward, roll the credits. ## The encore is the real test But here is the part that interests me more than the first title did. They went and did it again. Anyone can have one good year. A hot run of form, a few calls that fall right, a tournament where the bounces go your way, and suddenly you are champions. The first title answers the question "can you win?" It does not answer the much harder one, the question that actually separates a fluke from a fact: "was that real?" Only the second title answers that, and it answers it for good. The first win tells you it was possible. The second tells you it was not luck. Everything I have learned about building says the second is the one that counts. I see the same shape in companies all the time. The first time something works, a launch, a product, a good quarter, the honest feeling underneath the celebration is relief, plus a small private worry that you might not manage it twice. The encore is what kills that worry. It is the difference between "we got lucky" and "we know how to do this." It is also much harder, because the first time nobody expected anything and the second time everybody does. RCB won their first title carrying eighteen years of failure. They won their second carrying the weight of being champions, which is its own burden. Handling both is the entire point. ## Kohli, and the case for staying You cannot tell this story without Kohli. He has been with RCB since the franchise began in 2008, through every one of those eighteen empty years, in an era when almost everyone else chases the better deal at the better team. He could have left a dozen times and won sooner somewhere else. He stayed. And in the 2026 final, at thirty-seven, he reached a fifty in twenty-five balls, the fastest of his long IPL life, and finished unbeaten. There is a version of loyalty that is just stubbornness and a version that is conviction, and the only way to tell them apart is the ending. Kohli's looked like stubbornness for eighteen years. Now it looks like the best decision he never stopped making. The reward for staying is that when it finally works, it is unambiguously yours, with nobody to share the credit and nobody to blame for the wait. I have a soft spot for that, having argued before that the people who stay and actually operate the thing earn a kind of credibility that the people who pass through never do. ## What the IPL became It is worth remembering how improbable the whole stage is. When the IPL launched in 2008, plenty of serious people treated it as a gimmick, a noisy, cash-soaked experiment that real cricket would tolerate for a season or two. Instead it rewired the sport. It proved a domestic T20 league could become the financial center of gravity for a global game, it changed how players are valued and paid and developed, and it became, for two months a year, the thing an entire country arranges its evenings around. Almost every T20 league in the world now is a copy of a model people were sure would not last. There is a lesson in that alone, about betting on the format everyone calls a fad. But mostly, in 2026, the IPL is just the stage on which one team finally stopped being a joke and then, more impressively, refused to become one again. In the oldest piece on this site, written nineteen years ago, I argued that you cannot judge a cricketer by a single innings. You cannot judge a team by a single trophy either. You judge them by the second one. Which is why RCB, of all teams on earth, finally have nothing left to prove. NJ Nikhil Jathar “Two titles in, I still flinch when RCB need ten an over. Eighteen years leaves a mark.” --- ## Rebellion on Mars: what frontier sci-fi taught me · Nikhil Jathar URL: https://www.nikhilj.com/rebellion-on-mars-the-fight-for-freedom-and-independence/ # Rebellion on Mars: what frontier sci-fi taught me The Mars rebellion is one of sci-fis oldest stories. Underneath the spaceships, it is a builders story about earning the right to set your own terms. Nikhil Jathar18 April 20232 min read I have always been drawn to stories about rebellions on distant frontiers, and the most enduring version is the one set on Mars: a small, scrappy colony deciding it no longer wants to be governed by a faraway Earth that does not understand its life. It is one of the oldest shapes in science fiction, and I do not think we keep telling it by accident. Underneath the spaceships, it is a story about something very human, and something I think about more than I probably should as a founder. The Mars rebellion is really a story about the moment a place becomes its own thing. A colony starts as an extension of home, dependent and obedient, and then slowly develops its own conditions, its own problems, its own people who were born there and owe the old world nothing. At some point the distance becomes more than physical. The story is about that point, and the courage it takes to claim it. ## Why the frontier keeps producing this story Frontiers create independence whether anyone plans it or not. The people who go to the hard, far place have to solve problems the home world has never faced, with tools the home world does not have, and that experience changes them. After a while, being governed by people who have never stood where you stand starts to feel absurd. The rebellion is not really about anger. It is about a gap in understanding that grows until it cannot be bridged from a distance. I find this idea genuinely moving, and not only in fiction. Anyone who has built something far from the established centers of power knows the feeling of being told how to do a thing by someone who has never done it. ## What it teaches me about building The Mars story is, quietly, a builder's story. It is about the people willing to go to the hard place, take on the real risk, and earn the right to set their own terms. The freedom in these stories is never handed over. It is built, slowly, through competence and survival, until independence is simply the honest description of what already exists. That sequence, do the hard thing well enough for long enough and the autonomy becomes undeniable, is one I believe in deeply. So I keep returning to these stories, spaceships and all. They are not really about Mars. They are about the oldest pattern there is: that the people who go somewhere difficult and make it work eventually, and rightly, want to govern themselves. It is a good thing to be reminded of, on Earth, in a much smaller way, on an ordinary Tuesday. NJ Nikhil Jathar “Freedom in these stories is never handed over. It is built, slowly, until it is simply undeniable.” --- ## Software supply chain security, plain terms · Nikhil Jathar URL: https://www.nikhilj.com/software-supply-chain-security-slsa-sigstore-in-toto/ # Software Supply Chain Security in Plain Terms: SLSA, Sigstore, and in-toto SLSA, Sigstore, and in-toto without the jargon: how to prove a deployed artifact is what it claims to be, and was built the way you expect. Nikhil Jathar2 July 20265 min read When people say “software supply chain security,” most engineers nod and quietly hope someone else is handling it. The phrase sounds like compliance theater. It is not. A supply chain attack means an adversary compromises something you depend on (a base image, a build step, a package, a CI runner) so that malicious code ends up in your production system without anyone writing it there on purpose. The famous incidents of recent years were exactly this shape. The good news is that the cloud native ecosystem has built practical, open tools to defend against it. Three names come up constantly: SLSA, Sigstore, and in-toto. They solve different parts of the same problem, and together they form a coherent story. Here is that story without the jargon. ## The core question: can you trust this artifact? Every artifact you deploy (a container image, a binary, a package) arrives with an implicit claim: “I am what you think I am, and I was built the way you expect.” Supply chain security is about making that claim explicit and verifiable. You want to answer three questions: - Who produced this, and can they prove it? (signing) - How was it built, and from what? (provenance and attestation) - What level of rigor did that build process meet? (a framework to measure against) Signing, attestation, and a framework. Sigstore, in-toto, and SLSA line up against those three questions. ## SLSA: the yardstick SLSA (Supply-chain Levels for Software Artifacts, pronounced “salsa”) is not a tool. It is a framework, a set of levels that describe how tamper resistant your build process is. Lower levels ask for basic provenance: a signed statement of how an artifact was built. Higher levels demand that builds run on hardened, isolated infrastructure where even the person who wrote the pipeline cannot quietly inject something into a specific build. The value of SLSA is that it turns a vague goal (“be more secure”) into a ladder you can climb and measure. You can say “our releases are SLSA level 3” and that sentence means something specific and checkable, rather than being a marketing adjective. ## Sigstore: signing without the key management nightmare Signing artifacts proves who produced them. The historical blocker was key management: nobody wanted to hold a long lived private signing key. Sigstore removes that blocker with keyless signing. Instead of a key you keep, you authenticate with an identity (an OIDC account or a CI workload identity), receive a short lived certificate bound to that identity, sign in that brief window with an ephemeral key, and record the signature in a public transparency log called Rekor. The result: no long lived secret to leak, verification tied to a specific identity, and a public audit trail. In practice you sign with `cosign sign` and verify with an explicit identity assertion: `cosign verify $IMAGE \ --certificate-identity=releases@example.com \ --certificate-oidc-issuer=https://accounts.google.com` That verification is a real gate. If anything other than your release identity signed the image, it fails. (One footnote from contributing to cosign: keyless has been the default since v2, so ignore any older guide that still tells you to set `COSIGN_EXPERIMENTAL`.) ## in-toto: proving the steps in between Signing tells you who produced the final artifact. But a supply chain is a sequence of steps: fetch source, run tests, build, package, push. in-toto, a CNCF graduated project, secures that whole chain. It lets you define the expected steps and who is allowed to perform each one, and it produces signed attestations at each step. At the end, you can verify that the artifact you are about to run actually went through every required step, performed by the authorized party, in the right order. Think of it as a chain of custody for software. Sigstore signs the box; in-toto proves the box went through every checkpoint it was supposed to, and nobody skipped a step or swapped the contents along the way. ## How they fit together These are layers, not competitors: - in-toto produces attestations about how an artifact was built (provenance and step by step custody). - Sigstore signs those attestations and the artifact, binding them to a verifiable identity and recording them publicly. - SLSA is the yardstick that says how rigorous the whole process needs to be to earn a given trust level. In a modern setup, your CI builds an image, generates SLSA provenance and in-toto attestations describing exactly how it was built, signs all of it keyless with Sigstore, and pushes it. At deploy time, a Kubernetes admission controller verifies the signature and the attestations before the workload is allowed to run. A malicious image with no valid provenance from your pipeline simply never starts. ## Why this belongs to everyone, not just security teams For a long time this felt like specialist work. It is becoming table stakes. The tooling is open source, the flows integrate with the cloud native systems teams already run (OIDC identities, ephemeral workloads, declarative admission policy), and regulators and customers are increasingly asking for provenance by default. If you want to start, start small. Sign one of your images keyless and verify it with an identity constraint. Generate provenance for one build and read it. Add one admission policy that requires a signature. Each step is independently useful, and together they turn “we hope our supply chain is clean” into “we can prove it.” ## Takeaways - Supply chain security answers three questions: who made this, how was it built, and how rigorous was that process. - SLSA is the framework (the yardstick), Sigstore does identity based signing without long lived keys, and in-toto attests to the steps in between. - They compose: in-toto attests, Sigstore signs, SLSA measures, and a Kubernetes admission controller enforces at deploy time. - Start small and real: sign one image, verify with an identity, add one policy. The tooling is open and cloud native friendly. NJ Nikhil Jathar “Hope is not a supply chain strategy.” --- ## The best leadership lesson I ever learned · Nikhil Jathar URL: https://www.nikhilj.com/the-best-leadership-lesson-i-learned/ # The best leadership lesson I ever learned I learned it by getting it badly wrong during a scary week. Your team does not mirror what you say. It mirrors how you are, and managing that is the job. Nikhil Jathar1 July 20253 min read The single most useful thing I have learned about leadership, I learned by getting it badly wrong. Early on, during a genuinely scary week at the company, I did what felt responsible. I told my team the truth about how worried I was, in detail, repeatedly. I thought I was being honest and transparent. What I was actually doing was handing my anxiety to fifteen people who then could not sleep either, and who started making the jumpy, defensive decisions that anxious people make. The crisis got worse, and a good chunk of that was my doing. The lesson that came out of that week has shaped how I lead more than anything in any book: your team does not mirror what you say. It mirrors how you are. And managing how you are is most of the job. ## Composure is contagious, and so is panic I used to think leadership was mostly about decisions and communication, the visible things. I now think those sit on top of something quieter and more important, which is the emotional state you bring into the room. People take their cue for how worried to be from the most senior person present, almost entirely unconsciously. If I walk in tense, the meeting is tense before I have said a word. If I walk in calm, problems that looked like emergencies become, simply, problems to solve. This is not about hiding the truth or performing a fake confidence. People see through that instantly, and it is its own kind of poison. It is about doing the work of actually being steady, so that the steadiness you project is real. My team never once mirrored what I said. They mirrored how I was, every single time, and I was usually the last to notice. ## So the work moved inward Once I understood this, a strange thing happened. The most important leadership work stopped being external and became internal. Before a hard conversation or a tense all-hands, the most useful preparation is not rehearsing the words. It is getting my own state right, so that the calm is genuine and the team can borrow it. This is why I am so protective of my own recovery and the hours that keep me clear-headed. A depleted leader is a leaking source of anxiety, and the team pays for it without knowing why. ## What I do differently now When something genuinely goes wrong now, my first job is not to communicate the problem. It is to become the calmest accurate person in the building. Not falsely reassuring, calm and honest at the same time, which is harder and worth it. Then I can tell the team exactly what is happening, and because I am steady, they hear a problem we will solve rather than a reason to panic. The same facts, delivered from a different state, land in a completely different place. It took me years and one bad week to learn that the most powerful thing I bring to my team is not my decisions or my words. It is my composure, or my lack of it, which they will quietly absorb and amplify no matter how carefully I phrase the email. Get the state right, and most of leadership takes care of itself. Get it wrong, and no amount of good communication can save you. NJ Nikhil Jathar “My team never once mirrored what I said. They mirrored how I was.” --- ## The deep-work ritual: how I structure my week · Nikhil Jathar URL: https://www.nikhilj.com/the-deep-work-ritual-how-i-structure-my-week-for-code-strategy-and-leadership/ # The deep-work ritual: how I structure my week I run engineering, strategy and leadership across several products. My only sane response is a deep-work ritual I defend with real stubbornness. Nikhil Jathar10 December 20253 min read For years I ran my week like an inbox. Whatever arrived loudest got my attention first, and the tiredness at the end of each day felt like proof of work. It took me an embarrassingly long time, and a few quietly missed opportunities, to understand that the most important thing I decide each week is not what I will do. It is what I will refuse to let near the hours that matter. I run engineering, strategy, and a fair amount of leadership across several products, so the demand on a given day is effectively infinite. The only sane response I have found is a ritual, defended with more stubbornness than the work usually seems to deserve. This is that ritual, and the reasoning underneath it. ## Two mornings, blocked before anything else exists Before a single meeting goes on my calendar, two mornings are gone. Blocked, immovable, reserved for the hardest thing I am responsible for that quarter. Not email. Not reviews. The one problem that, if it goes well, makes most of the other noise irrelevant. The order matters more than the amount. If I protect the mornings first and let everything else fill in around them, I get my best work and a slightly annoyed calendar. If I do it the other way, the mornings never come, because there is always something with a louder claim. There is always something louder. That is precisely why it cannot be allowed to vote. ## Deep work is fragile for the first hour The reason I am so severe about this is that hard thinking does not switch on. When I sit down with a genuinely difficult problem, the first twenty minutes are useless and the work only becomes load-bearing after an hour. So an interruption does not cost a minute. It costs the hour I had not yet earned, and there are only so many of those in a week. An interruption never costs you a minute. It costs you the hour you had not finished earning. ## The unglamorous mechanics The system itself is dull, which is how I know it works. The hard thing is single-threaded, one problem at a time, because context-switching between two difficult things means doing neither well while feeling busy. The shallow things, the reviews and replies and approvals, get batched into the afternoons, where shallow things belong. And a meeting, in my book, is a decision that failed to happen in writing first, so I treat every meeting request as a small accusation that something was not thought through. This connects to how I handle the rest of the flood, which I have written about under the idea of an intentional inbox, and to why I am deliberate about protecting the recovery time that makes the deep hours possible at all. None of it is clever. All of it is just defended. ## The week I was proudest of looked like nothing The best week I had this year would have looked, to most people, like I did very little. It had two long quiet mornings, a great deal of reading, and one decision I had been avoiding for a month. Everything good that followed traces back to those hours I very nearly gave away to something urgent and forgettable. The discipline is not in the planning. Anyone can draw a perfect week on a Sunday. The discipline is in the defending, on a Wednesday, when someone reasonable wants one of your mornings and you have to decide whether you meant any of it. NJ Nikhil Jathar “I protect two mornings a week. The other five fight over the scraps.” --- ## The four questions I got after the IRI talk · Nikhil Jathar URL: https://www.nikhilj.com/the-four-questions-i-got-after-the-iri-talk/ # The four questions I got after the IRI talk Four questions came back after my IEEE IRI 2026 talk: why not just use ChatGPT, how invariants differ from unit tests, and whether it is really reuse. Nikhil Jathar8 August 20266 min read On 31 July I presented a paper at IEEE IRI 2026, the conference on information reuse and integration. The paper is an industry-track case study of an accounting system where an AI agent writes to the general ledger, and the talk went the way talks go: twenty minutes of me, then the part that actually matters. Four questions came back. I want to write them down while the answers are still the ones I gave in the room rather than the ones I polished afterwards, because the polished version is always less useful. ## 1. Why not just use ChatGPT or Claude directly? This one comes first every time, and it is a fair question rather than a naive one. If a model can read a bank statement and describe the journal entry, why build anything underneath it? Here is the difference, stated as plainly as I can. Ask a chatbot to post a journal entry and you get text that looks like a journal entry. Ask our system and you get a row in an immutable ledger that provably balances, because deterministic code wrote it and a set of structural checks verified it before it landed. A bare model has no system of record. It has no invariants. It has no ownership model that says which component is allowed to write where. And it cannot be unit tested, because you cannot unit test a prompt. You can evaluate a prompt, which is a statistical claim about a distribution of outputs. That is a different thing from a test, and an auditor knows the difference even if they would not put it that way. What the model is genuinely good at is deciding what you meant. That is the hard, unstructured half of accounting, and handing it to a model is the actual innovation. It just never decides what the books say. ## 2. How is an invariant different from a unit test? You already have thousands of tests. The best way I have found to answer this is with a hundred dollars. You transfer 100 from savings to checking. The unit test asks: did checking go up by 100? It did. The test passes, green tick, ship it. But the code forgot the other half. Savings never went down. You have created 100 out of nothing, with every test green, because the test only checked what it was told to check. An invariant asks a different question, and it asks it after every ledger operation rather than once per test case: do all debits still equal all credits across the entire book? That fails immediately. It does not care which function ran or what the author of that function thought they were doing. The line I keep coming back to: unit tests check that the code did what you asked. Invariants check that the books still make sense. The paper reports five bugs that unit tests passed and invariants caught. That is not five sloppy functions. Each of those five times the function did exactly what it was asked to do and left the ledger broken anyway, which is precisely the class of bug a test suite written by the same person who wrote the code will never find. ## 3. The model is nondeterministic. How can it be anywhere near a general ledger? It is not. That is the whole answer, and the rest is showing where the boundary sits. The model interprets. Deterministic code writes. Invariants verify. Nondeterminism is real and it does not go away because you disapprove of it, so the design question is where you let it land. Interpretation is a good place: if the model misreads an ambiguous line item, that is visible, correctable, and cheap. The ledger write is a terrible place, so it never happens there. The second half of the answer is the ownership rule, which is the part the IRI audience found more interesting than I expected. Any module can read any table. Each table has exactly one module allowed to write to it. So when a sales invoice is wrong there is exactly one place to look, one file, not the entire system. It also contains the blast radius: a defect in one industry module cannot corrupt the sales ledger, because it never had write access to it in the first place. The alternative, which most systems arrive at without ever choosing it, is a document that eleven people can edit, where fixing a defect starts with reading the version history and guessing. ## 4. Is that actually reuse, or is it configuration with better marketing? The sharpest question of the four, and the right one to ask at a conference about reuse. My answer is the chef and the recipe. A trained chef already knows knife work, heat, and timing. Hand them a recipe for a dish they have never cooked and they will make it tonight, because the recipe carries what is specific and the chef carries what is general. The model is the chef. The module manifest is the recipe. The shared database is the kitchen. Adding an industry means writing one more recipe, not sending the chef back to culinary school. The concrete instance in the paper is a generated module for a tattoo parlour: 8 tables, 18 actions, 48 of 48 tests passing. Across three generated modules there were 19 human decisions in total, and every one of them was a business preference rather than an accounting decision. How long a deposit is held, for example. Nobody had to decide how double entry works, because that is in the chef. So is it reuse? What gets reused is the interface contract and the accounting core. What varies is a manifest. I think that is a stronger claim than configuration, because configuration implies switches on a fixed system, and this generates a module that did not exist. But I would rather argue about the label than lose the distinction, which is that conventional ERP answers a new industry by hiring a second chef and building a second kitchen. ## The question I expected and did not get Nobody asked whether this generalises outside accounting. I would have struggled with it, honestly. Accounting has unusually crisp invariants: debits equal credits, and the whole discipline has spent five centuries agreeing on what correct means. Not every domain has a rule that clean, and the parts of this architecture that depend on having one are the parts I would be most careful about promising elsewhere. If you were in the room and asked something I have mangled here, tell me and I will correct it. NJ Nikhil Jathar “The model decides what you meant. It never decides what the books say.” --- ## Future of work, part 2: human creativity · Nikhil Jathar URL: https://www.nikhilj.com/the-future-of-work-embracing-human-creativity-in-an-ai-dominated-world-part-2/ # The future of work, part 2: human creativity in an AI world As AI gets better at the mechanical parts of work, it does not make human creativity less valuable. It quietly makes it the main thing worth paying for. Nikhil Jathar5 November 20242 min read In the first part of this, I argued that the AI shift in how we work is real, and that both the panic and the hype around it are mostly noise. This is the more hopeful half of the argument, and the one I believe more strongly the longer I run a company through this change. As AI gets better at the mechanical parts of work, it does not make human creativity less valuable. It quietly makes it the main thing left worth paying for. I do not mean that as a comforting slogan to calm nervous people. I mean it as something I watch happen in my own company, week after week, as the floor of what software can do for us keeps rising. ## What rises in value when the routine gets cheap When a capable model can handle the routine version of a task, the routine version stops being where anyone earns their keep. What becomes scarce, and therefore valuable, is everything the model cannot do: deciding which problem is worth solving in the first place, having the taste to know when an answer is merely correct rather than good, and the judgment to be accountable for a decision when it goes wrong. These were always the senior parts of any job. AI is just removing the busywork that used to hide how much they mattered. So the future of work, as I see it from inside a company living through it, is not humans competing with machines on the machine's turf. It is humans being pushed, sometimes uncomfortably, up the stack toward the work that was always the actual point. AI did not make human creativity less valuable. It quietly removed everything around it, until creativity was the only thing left to pay for. ## The uncomfortable part I will not pretend this transition is painless. Being pushed up the stack is only good news if you can make the climb, and not everyone is given the time or the support to. The honest version of this argument includes the people for whom the routine work was not busywork but their whole role, and a company that cares about its people has to take that seriously rather than wave it away with a slogan about creativity. But the direction is clear, and it is more human than the doom narrative suggests. The skills that hold their value are the deeply human ones: imagination, taste, judgment, the ability to care about the right things. That is what I try to protect and prioritize as the tools get more capable. ## Why I am, on balance, optimistic Every generation invents a tool it is sure will hollow out work, and every time, the tool ends up changing what work is rather than ending it. I think this is another turn of that very old wheel. The mechanical parts get automated, the human parts become more central, and the people who thrive are the ones who lean into exactly what makes them human rather than trying to out-machine the machine. That is not a threat. For anyone willing to do the harder, more creative work, it is close to the best news in a generation. NJ Nikhil Jathar “AI did not make human creativity less valuable. It quietly made it the only thing left worth paying for.” --- ## The future of work, part 1: navigating a new reality · Nikhil Jathar URL: https://www.nikhilj.com/the-future-of-work/ # The future of work, part 1: navigating a new reality The future of work is not the vendor utopia or the headline apocalypse. It is a new reality to navigate carefully, by the people actually living it. Nikhil Jathar24 September 20242 min read Every few years a technology arrives that people are certain will end work as we know it, and every time, the reality turns out to be both less dramatic and more disruptive than either the optimists or the doomers predicted. We are living through one of those moments with AI, and the loudest voices on both sides are, in my experience, the least useful guides to it. The future of work is not the utopia the vendors sell or the apocalypse the headlines threaten. It is a new reality that has to be navigated carefully, by people actually living in it. I run a company through this change rather than commentate on it, so this is the view from inside the thing, where it is messier and more ordinary than the takes suggest. ## The shift is real, the framing is wrong Something genuine is happening. The cost of a whole category of work is falling fast, and pretending otherwise would be foolish. But the popular framing, humans versus machines in a fight one side wins, gets the shape of it wrong. In practice it looks less like replacement and more like a steady redrawing of the line between what we hand to the tool and what we keep for ourselves. That line has moved many times before. It is moving again, faster this time. The mistake the panic makes is treating today's snapshot as the destination. The mistake the hype makes is the same, just more cheerful. Both freeze a moving thing and call it the future. Every generation believes its tools will end work. So far, the tools have only ever changed what the work is. ## What navigating it actually requires Navigating a new reality is not a matter of grand predictions. It is a series of small, practical, slightly boring choices, made repeatedly. Which tasks do we genuinely hand over, and which do we keep because the judgment matters too much. How do we help people move toward the work that is becoming more valuable rather than leaving them stranded on the work that is becoming cheap. What do we refuse to automate because doing so would quietly remove the human accountability we actually need. These are not visionary questions. They are operational ones, and they are the real substance of the future of work, far more than any keynote about it. We work them out by living them, not by forecasting them. ## Why I stay calm about it I am neither frightened nor giddy about all of this, and I think that middle position is the honest one. The new reality is disruptive and it is navigable, both at once. The people and companies that do well will not be the ones with the best prediction about where it all lands. They will be the ones who stay adaptable, keep their judgment sharp, and treat the change as something to be steered rather than either feared or worshipped. In the second part, I make the more hopeful case for why the deeply human skills are exactly the ones that rise in value. The short version is that they always have. NJ Nikhil Jathar “Every generation believes its tools will end work. So far, the tools have only ever changed what the work is.” --- ## The read-to-lead habit: five books · Nikhil Jathar URL: https://www.nikhilj.com/the-read-to-lead-habit-5-books-that-arent-about-business-but-made-me-a-better-leader/ # The read-to-lead habit: five books that made me a better leader The worst leadership advice I ever read came from a leadership book. Why the books that actually made me a better leader were never about business at all. Nikhil Jathar20 May 20252 min read The worst leadership advice I have ever read came from a leadership book. It was confident, well-structured, and completely useless in the one situation I bought it to handle. Over time I noticed that almost everything I actually use as a leader, I picked up somewhere else entirely, usually from a book that had nothing to do with business at all. So I stopped reading books about leading and started reading more or less everything else, and I became a better leader for it. I want to be honest that your list will look nothing like mine, because the point is not the specific titles. The point is the angle. The books that taught me the most about running a company were never trying to teach me anything about running a company, which is exactly why the lessons stuck. ## Why business books mostly fail A business book has a problem built into its premise. It has to flatter the reader, justify its own thesis, and pretend the messy thing it describes is more orderly than it is. So you get a tidy framework, three case studies chosen because they fit, and a confidence the real world rarely supports. You finish it feeling informed and act no differently on Monday. A great novel does the opposite. It refuses to simplify people, and leading people is most of the job. ## The five that did it for me If you forced me to name them: Marcus Aurelius on staying composed when everything is on fire, which is most weeks. A serious history of a hard decade, because nothing teaches decision-making under uncertainty like watching real people make irreversible calls without knowing how it ends. A novel that made me feel the inside of a life unlike mine, which is the closest thing to a superpower a manager can have. A book about how science actually gets made, all the wrong turns the textbooks edit out. And one quiet book about attention, because protecting it turned out to be the whole game. You should swap every one of those for your own. The category matters more than the title. The best leadership ideas I have rarely arrived dressed as leadership ideas. They came in through the side door, from a book about something else. ## What outside reading actually trains Reading widely does something a business book cannot. It builds the muscle for holding a complicated situation in your head without rushing to flatten it into a framework. It grows the empathy that leading people actually requires. And it keeps reminding you that the smartest people in history were usually wrong about something, which is a useful thing to carry into a room where everyone expects you to be certain. I protect my reading time the way I protect the rest of my recovery, because it is recovery, and it is also the most reliable source of ideas I have. The habit is simple. Read mostly outside your field, read things that are better than you, and trust that the lessons will find their way to the right problem on their own. They reliably do. NJ Nikhil Jathar “The worst leadership advice I ever read came from a leadership book.” --- ## The social-media dopamine dilemma · Nikhil Jathar URL: https://www.nikhilj.com/the-social-media-dopamine-dilemma-how-social-media-affects-our-brain-chemistry/ # The social-media dopamine dilemma Social apps are engineered to hijack your attention with the same chemistry as a slot machine. Why willpower is the wrong tool, and what actually helps. Nikhil Jathar11 January 20242 min read I think about attention for a living, in the sense that protecting mine is most of how I get any real work done. So the way social media is engineered to hijack it is not an abstract concern for me. It is a daily fight, and one I do not always win. I wanted to write down what I actually understand about the dopamine machinery underneath it, without the moral panic and without pretending I am above it. The short version is that these apps are not badly designed. They are designed extremely well, for a goal that is not mine. Every pull-to-refresh, every unpredictable reward, every red dot is tuned to trigger a small dopamine response, the same chemistry that makes slot machines work. The genius and the problem are the same thing. It is variable reward, delivered just often enough to keep you checking, and your brain was never built to resist it. ## Why willpower is the wrong tool For a long time I treated my phone habit as a failure of discipline, which made me feel worse and changed nothing. The reframe that helped was understanding that I was not fighting my own weakness. I was fighting teams of brilliant people whose entire job is to win exactly this fight against me, with far better data than I have about myself. Willpower against that is a losing strategy. It is like trying to out-stubborn a casino. So I stopped relying on willpower and started changing the environment instead, which is the only thing that has ever actually worked. ## What actually helps The same principle runs through how I protect my deep-work hours and run an intentional inbox. You do not win the dopamine fight in the moment. You win it earlier, by design. Notifications off by default. The apps off the home screen, or off the phone. Friction added deliberately where the apps removed it. None of this is heroic. It just moves the battle from my exhausted in-the-moment self to my calmer earlier self, who is much better at this. I am not anti-technology, obviously. I build it. But I have a particular distaste for technology designed to exploit a vulnerability rather than serve a need, and the dopamine economy is the clearest example we have. The healthiest thing I have done is stop blaming myself for losing a fight that was rigged, and start quietly refusing to show up to it on the house's terms. NJ Nikhil Jathar “I stopped blaming myself for losing a fight that was rigged, and started refusing to show up on the houses terms.” --- ## Five underrated comedy TV shows worth finding · Nikhil Jathar URL: https://www.nikhilj.com/top-5-underrated-comedy-tv-shows/ # Five underrated comedy TV shows worth finding The comedies that were brilliant and never got their due. Why great shows get overlooked, and the quiet joy of rescuing one for a friend. Nikhil Jathar19 December 20222 min read Everyone knows the famous comedies. The ones that get quoted at parties, that show up on every list, that you are vaguely embarrassed not to have seen. I am more interested in the other category: the comedies that were genuinely brilliant and somehow never got their due, the ones I find myself recommending again and again to people who have never heard of them. Here is a quiet defense of the underrated. An underrated comedy is a specific thing. It is not a show that failed because it was bad. It is a show that was excellent and, for reasons of timing or marketing or sheer bad luck, never found the audience it deserved. Television is full of these, and tracking them down is one of my small, genuine pleasures. ## Why great comedies get overlooked The pattern is usually one of a few things. The show was ahead of its time, doing something the audience was not ready for until years later. Or it was badly scheduled and quietly buried by a network that did not understand it. Or it was simply too odd, too specific, too unwilling to soften itself into something easily marketed. The very qualities that make these shows special are often the same ones that kept them from being popular, which is a slightly heartbreaking irony. The best of them have a confidence about their own strangeness. They were clearly made by people who cared more about being good than about being liked, and you can feel it. ## Why I keep hunting for them There is a particular joy in discovering something brilliant that the world overlooked, and an even better one in pressing it on a friend and watching them get it. It feels less like consuming and more like rescuing something. As someone who has spent a career building things that had to fight to be noticed, I have a real soft spot for the excellent work that did not get its moment. Quality and recognition are not the same thing, in television or anywhere else, and the gap between them is where a lot of the most interesting work lives. So my recommendation is to go looking past the famous names. Some of the funniest television ever made is sitting quietly in the places the algorithms do not push, waiting for someone to notice it was great all along. Finding it is half the fun. NJ Nikhil Jathar “Quality and recognition are not the same thing, in television or anywhere else.” --- ## The funniest TV comedies since 2000 · Nikhil Jathar URL: https://www.nikhilj.com/top-funny-tv-shows-post-2000-laugh-out-loud/ # The funniest TV comedies since 2000 I switch off with comedy that is actually funny, not clever or important. My opinionated take on the best TV comedy since 2000, and why funny is so hard. Nikhil Jathar16 April 20242 min read I spend most of my working life thinking about software, so when I switch off, I do not want anything that asks me to think. I want to laugh. Over the years I have become quietly opinionated about what counts as genuinely funny television since 2000, and since this is my site and nobody can stop me, here is the short version of my argument. My one rule is that a comedy has to actually be funny, not merely clever or heartwarming or important. A lot of what gets called the best comedy of the century is really very good drama with jokes in it. Fine, but that is a different category, and I am here for the shows that make me laugh out loud alone on a sofa, which is the only honest test. ## The ones that pass the sofa test The American workplace comedies of the 2000s mostly hold up, because awkwardness ages well and nobody has ever stopped finding a boss who tries too hard funny. The British half-hours, shorter and meaner, hold up even better, because they end before they fall in love with their own characters. And the animated comedies quietly became the bravest writing on television, because a cartoon can say the thing a live actor cannot. I notice I keep returning to the shows that are confident enough to be a little uncomfortable. Comfortable comedy is just background noise. The funny ones take a small risk in almost every scene. ## Why I think comedy is underrated as a craft Here is the founder in me sneaking back in. Making something genuinely funny is one of the hardest creative problems there is, far harder than making something serious, because a joke either works or it does not and the audience tells you instantly. There is no hiding behind importance. I have a real respect for that kind of feedback loop, the brutal, immediate kind, which is probably why I admire great comedy writers more than almost any other artists. So no grand conclusion here, just a recommendation. The next time you have had a long week, skip the prestige drama everyone says you should watch and put on something that was built, with enormous skill, to do nothing more noble than make you laugh. It is harder to make than it looks, and on a Friday it is worth more than almost anything more serious. NJ Nikhil Jathar “Making something genuinely funny is harder than making something serious. The audience tells you instantly.” --- ## What 7 days of AI-assisted coding changed · Nikhil Jathar URL: https://www.nikhilj.com/what-seven-days-of-ai-assisted-coding-changed/ # What seven days of AI-assisted coding changed about how I build I built a server tool in a week with an AI pair. The build was the easy part. What it permanently changed about how I scope, delegate, and review. Nikhil Jathar17 March 20264 min read A little over a year ago, I built a working server tool in about a week with an AI model as my pair. People still ask me about the speed, as if seven days were the interesting part. They were not. The build was the easy part. The part worth talking about is what those seven days quietly changed in how I work now, on everything, long after the novelty wore off. I want to be precise about this, because there is a lot of noise on both sides. AI-assisted coding did not make me a ten times engineer, and it did not make engineering obsolete. It moved the work. It took the part I was good at and made it cheap, and it took the part I used to skip and made it the whole job. ## The build was the easy part For seven days, the model wrote most of the lines. I described what I wanted, it produced a first draft, and I corrected it. The boilerplate, the glue code, the second and third versions of a function, all of it came fast. If you measured the week in lines of code per hour, it looked like a miracle. But lines of code per hour is a deeply silly way to measure building, and it always was. The miracle wore off the moment the tool met a real server with a real edge case, and I had to understand every line I had not actually written. Which, it turns out, is a different skill from writing it yourself. ## From writing code to specifying intent The biggest shift is that my job is now mostly upstream of the code. I spend my time saying, as exactly as I can, what the thing should do and why. The clearer I am about intent, the better the output. The vaguer I am, the more confidently the model gives me something plausible and wrong. This sounds like a small change. It is not. For twenty years, the act of writing the code was also the act of thinking through the problem. The typing was slow enough that it forced clarity. Take the typing away and you can produce a great deal of code before you have actually finished thinking, and the model will happily help you do exactly that. So I had to learn to do the thinking on purpose, separately, before I let the machine type. AI did not make me twice as fast. It made me twice as fast and exactly as wrong, until I learned to slow the thinking down. ## The debugging tax nobody mentions Here is the part the demos skip. Code you did not write is code you do not yet understand, and understanding it later, under pressure, costs time. The first week I leaned hard on AI output, I shipped quicker and then spent the following week paying it all back in confused debugging, because I was now maintaining decisions I had never consciously made. So I changed the rule. The model can draft anything. But nothing ships until I can explain every meaningful line as if I had written it, because in the way that matters, I am about to own it. Our testing tooling catches a lot, but a test only checks the behavior you thought to check. Judgment is still mine. ## What now counts as a weekend project The practical effect is that the line between "too small to bother" and "too big to attempt" has moved. Things that used to need a sprint now genuinely fit a weekend, so we attempt more of them. That is the real productivity gain, and it is large. But the same shift makes it dangerously easy to start things, because starting is now almost free. Finishing, maintaining, and operating them costs exactly what it always did. This is why I am stricter than ever about asking why before I start. When building gets cheap, the scarce discipline is not building. It is choosing. ## The part that did not change After all of it, the things that decide whether software is any good are exactly the things the model cannot do for me. Picking the right problem. Having taste about the shape of a solution. Knowing which 5 percent of the work is the actual point. Being accountable when it breaks. The model is a remarkable typist and a fast, confident, occasionally hallucinating junior. I am still the one who has to run the thing in production and answer for it. So I have stopped asking whether AI will replace engineers. The better question is which part of engineering you were actually being paid for. If it was the typing, that part is now cheap. If it was the judgment, you just got more valuable, and a lot busier. NJ Nikhil Jathar “The machine wrote the code. I still own the 2am pager.” --- ## What we chose to patent: ERPClaw, TailTest · Nikhil Jathar URL: https://www.nikhilj.com/what-we-chose-to-patent-erpclaw-and-tailtest/ # What we chose to patent, and why: inside ERPClaw and TailTest We filed patents on two products and not the rest. What was novel enough to protect, how a small studio thinks about IP, and what a patent really buys you. Nikhil Jathar1 June 20263 min read We recently filed patents on two of our products, ERPClaw and TailTest. I want to write about that honestly, because most companies either say nothing about their patents or wave them around like trophies, and both are a little dishonest. A patent is not a trophy. It is a bet, with lawyers, and the only interesting question is why you chose to make it on one thing and not another. A small studio has no business filing patents on reflex. They are slow and expensive, and for most software the protection is thinner than founders imagine. So our default is to not bother. The two times we did bother are therefore the interesting cases, because they tell you what we actually think is novel rather than merely good. ## The filter: novel, not just nice Plenty of what we build is good. Very little of it is novel in the sense that matters for a patent, which is a specific mechanism that is genuinely non-obvious and that we would mind a competitor copying wholesale. Most software is a smart arrangement of known parts, and there is no shame in that. The filter I used was simple. Would I be annoyed in a deep, structural way if a well-funded competitor lifted this exact idea? For almost everything, the honest answer is no, because the moat is execution and operation, not the idea. Twice, the answer was yes. ## ERPClaw: making an ERP answer back A traditional ERP is a vast, rigid system that makes you learn its language. ERPClaw inverts that. You ask it a question the way you would ask a competent colleague, and it does the unglamorous work of translating your intent into the rigid world underneath, then translating the answer back. The novel part is not the chatbot. Chatbots are a commodity. The part worth protecting is the specific mechanism that maps loose human intent onto a strict enterprise data model reliably enough to trust with real operations. That mapping is hard, it is non-obvious, and it is the thing I would mind seeing copied. We did not patent the idea of an idea. We patented the specific, awkward mechanism that makes the idea survive contact with a real system. ## TailTest: testing that thinks like a skeptic TailTest came out of an old frustration. Testing tools check the behavior you remembered to specify, which means they are blind to exactly the failures you did not think of, which are the ones that hurt. The mechanism we filed on is about generating and prioritizing tests the way a suspicious senior engineer would, going after the cases a tired human would skip. As AI writes more of our code, this matters more, not less. When a model produces code at speed, the scarce thing is judgment about what could go wrong, and that is precisely what we tried to bottle. ## What a patent actually buys a startup Honestly, less than founders hope and more than cynics claim. It does not stop a determined competitor; it raises their cost and slows them down. It does not build a business; a patent on a product nobody wants is an expensive certificate. What it does give a small company is a small, specific edge in a world tilted toward whoever has more capital. We did not file to look impressive. We filed on two mechanisms we run in our own production every day, which is the only kind of innovation I trust, because we build and run the AI we recommend. If there is a lesson, it is that the patentable thing and the impressive thing are rarely the same thing. The impressive thing is usually the whole product. The patentable thing is one stubborn, specific mechanism buried inside it, the part you bled over and would hate to see lifted. Find that, and you know what you actually invented. NJ Nikhil Jathar “We patented two things and named neither of them after ourselves. Growth.” --- ## Why I ask what is the why before any new project · Nikhil Jathar URL: https://www.nikhilj.com/why-i-ask-whats-the-why-before-starting-any-new-project/ # Why I ask what is the why before any new project One question has killed more of my bad ideas than any plan or strategy session. Why I make every new project answer it clearly before we start building. Nikhil Jathar30 March 20252 min read There is one question I ask before starting anything new, and it has killed more of my bad ideas than every plan, spreadsheet, and strategy session I have ever run. The question is simply: what is the why. Not what will we build, not how, not when. Why does this deserve to exist, and why us, and why now. It sounds almost too simple to be useful. It is the most useful habit I have. I came to it the way most people come to good habits, by repeatedly suffering the absence of it. I have started projects that were technically interesting, commercially plausible, and completely pointless, and the common thread was always the same. Nobody had made me answer the why out loud before we began, so we discovered the emptiness three months in, expensively. ## The why is a filter, not a formality Most teams treat purpose as a box to tick, a line at the top of a document everyone skips. I treat it as a gate. If we cannot give a clear, specific, slightly uncomfortable answer to why this project should exist, we do not start it, no matter how exciting the how is. An exciting how attached to a weak why is the single most expensive thing a team can chase, because it feels like progress the entire time it is wasting your months. A good why has a particular quality. It is specific enough to argue with. "Customers want this" is not a why, it is a hope. "This specific customer keeps hitting this specific wall and pays us anyway" is a why you can build on. An exciting how attached to a weak why is the most expensive thing a team can chase. It feels like progress the whole time it is costing you. ## Why it matters more with a small team When you run several products with a small team, the why is not a philosophical nicety. It is survival. With fifteen people, every project we start is several projects we cannot, so an unjustified yes is really a hidden no to something better. The why is how I keep us from drifting into work that is busy rather than important, which is the quiet way small teams fail. This is the same discipline that runs through how I prioritize across everything. Choosing is the job. ## Asking it of myself, too The hardest place to apply the question is inward, on the projects I personally want to do because they are fun or flattering. The why is uncomfortable precisely because it does not care what I find interesting. More than once it has talked me out of something I was excited about, and every single time, with a bit of distance, it was right. A clear no in week one is a gift. A muddy yes that unravels in month four is the thing the question exists to prevent. So I keep asking it, on everything, even when it is annoying, especially when it is annoying. The discipline is not in having a good answer. It is in being willing to not start when the honest answer is that there isn't one. NJ Nikhil Jathar “Most of my best decisions were really just good questions I asked early enough.” --- ## Always Sunny in Philadelphia: best comedy · Nikhil Jathar URL: https://www.nikhilj.com/why-its-always-sunny-in-philadelphia-is-tvs-best-comedy/ # Why Its Always Sunny in Philadelphia is TVs best comedy An extreme position, defended sincerely: Its Always Sunny is the best comedy on television, and the cruelty that puts people off is exactly the point. Nikhil Jathar16 April 20242 min read I will defend a fairly extreme position: It's Always Sunny in Philadelphia is the best comedy on television, and it is not particularly close. People look at me strangely when I say this, usually people who watched two episodes, found the characters horrible, and switched off. That reaction is the whole point, and missing it is missing the joke entirely. Most comedies want you to love their characters. Sunny wants you to understand, slowly and with growing horror, that its characters are terrible people, and then it makes their terribleness so precisely observed that you cannot look away. That is a much harder and braver thing to pull off, and they have pulled it off for longer than almost any live-action comedy in history. ## Why the cruelty works The trick is that the show is never on the side of its characters. It is always, quietly, on the side of the audience watching them fail. The Gang are selfish, deluded, and incapable of growth, and the show knows it completely, which is what separates it from the lazy mean-spirited comedy it could easily have become. The cruelty has a point of view. You are laughing at delusion, not at a target. It also commits. Most shows soften over time, reaching for warmth and the audience's affection. Sunny has spent its entire run refusing to let its characters become lovable, which takes a discipline I genuinely admire as someone who knows how tempting it is to chase approval. ## The craft underneath the chaos What looks like anarchy is, structurally, incredibly tight. The plots are little farce machines where everyone's worst instincts collide at exactly the right moment. That precision is invisible because the surface is so chaotic, which is the highest form of craft, the kind that hides itself completely. The best engineering looks effortless too, and is never effortless. I respect anything built that well that pretends it was thrown together. So if you bounced off it early, I understand, but I would gently suggest you missed the design. Give it a proper run and watch for how careful the supposedly careless thing actually is. It is the funniest argument I know for the idea that the hardest craft is the kind nobody notices. NJ Nikhil Jathar “Most comedies want you to love their characters. Sunny wants you to understand they are terrible, and that is the harder trick.” --- ## Why we build and run the AI we recommend · Nikhil Jathar URL: https://www.nikhilj.com/why-we-build-and-run-the-ai-we-recommend/ # Why we build and run the AI we recommend I am a founder who ships, not an advisor with slides. Why AvanSaber runs the AI it sells, and what operating our own software actually teaches us. Nikhil Jathar18 February 20265 min read Early in my career, I sat through a lot of meetings where people sold software they had never run. The decks were always excellent. Clean diagrams, arrows pointing the right way, and a presenter who had quite obviously never been woken at two in the morning because a job queue had jammed and a customer's invoices were stuck somewhere between two systems. I learned a great deal in those rooms. Mostly, I learned what I did not want to become. That is the whole reason AvanSaber works the way it does, and the reason this site exists at all. We build and run the AI we recommend. We do not advise from a safe distance. We ship our own products, operate them in production, and let the bruises do the teaching. It sounds like a strategy when I put it on a slide. Most days it feels more like a refusal to bluff. ## The advice nobody has to live with I spent the first stretch of my career in enterprise ERP, a world where the gap between advice and consequence is enormous. A consultant recommends an architecture, signs off, and moves to the next account. Two years later, someone in a billing department is quietly building spreadsheets to paper over the cracks in a system that looked flawless in the proposal. I am not being cynical about consulting. A lot of that advice was genuinely good. The problem is structural. When you never have to operate the thing you designed, you slowly stop noticing the difference between an idea that is correct and an idea that is true. Correct is the architecture diagram. True is the 2am page. They are not the same thing, and only one of them improves your judgment. ## So we run our own stack Today we ship across an uncomfortably wide range. ERPClaw puts a conversational AI layer over the kind of ERP I used to implement by hand. TailTest tests software the way an impatient senior engineer would, if the senior engineer never got tired. There is SiteKit, a server tool I put together in about a week because I was tired of paying a subscription for something I could run myself. Inventory, SEO, social, payments. Frankly, more than a sensible person would attempt with the team we have. Every one of those products pages someone on our team when it breaks, and often that someone is me. This is the point, not an accident of being short-staffed. When the assistant inside ERPClaw gives a confidently wrong answer to a real customer, I do not read about it later in a satisfaction survey. I feel it that evening. That feeling is the most valuable input we have, and there is no way to buy it, outsource it, or fake it in a demo. ## A portfolio is a forcing function People ask why we run so many products with a team of fifteen. The honest answer is that breadth keeps us honest. My co-founder Varun and I split the company cleanly. He runs the business, I run the building, and between us we keep a standing argument going about how much is too much. He is usually right. But a portfolio does something a single product cannot. It forces you to build things that are genuinely reusable, because you simply do not have the people to solve the same problem twice. I trust a piece of software the way I trust a restaurant. Only if the people who made it actually eat there. So one deployment path runs under all of it. One design system. One testing harness, which happens to be a product we also sell. When I say that I ask what is the why before starting anything new, this is what it means in practice. With fifteen people and eight products, an unjustified project is not a small mistake. It is a month that nobody on the team gets back. ## What a customer actually gets from this A vendor who runs their own software is a different kind of vendor. We have usually already hit the failure you are about to hit. We have already argued about the awkward edge case at a whiteboard. So when we tell you an AI feature is ready, we are really telling you it survived contact with our own production, our own customers, and our own irritation. That is a narrower promise than most AI companies are making right now. It is also a more honest one. It keeps our enthusiasm in check too, which matters more than it sounds. It is easy to love a model in a keynote. It is much harder to stay in love with it after you have had to support it for six months. The few technologies I am still genuinely excited about are the ones that survived the second feeling, not just the first. ## The limit I keep running into Running what you build is clarifying, but it is not a magic trick. It does not tell you which product to kill, only that the killing is overdue. It does not make the hard call about a feature customers love and we cannot maintain. And it carries a real cost, which is focus. When you operate everything, everything has a claim on your attention, and attention is the one resource a small team cannot manufacture. I am strict about defending the hours that matter precisely because the build-and-run instinct quietly tries to eat all of them. So we draw lines. Not everything deserves to be built, and not everything we built deserves to keep running. The discipline is not in shipping more. It is in being able to look at something you are proud of and admit that it has had its time. None of this makes us right about where AI is going. It only means that when we are wrong, we find out first, and we find out for real, on a Tuesday, with a customer on the other end. For a company selling AI in 2026, that might be the only credential worth having. NJ Nikhil Jathar “Opinions here are mine. The bugs, unfortunately, are also mine.” --- ## Work-life integration vs. balance: my philosophy · Nikhil Jathar URL: https://www.nikhilj.com/work-life-integration-vs-balance-a-serial-entrepreneurs-philosophy/ # Work-life integration vs. balance: my philosophy Balance assumes two opposing weights and a virtuous person keeping them level. After a decade of running companies, I stopped believing in the scale. Nikhil Jathar22 October 20253 min read I get asked about work-life balance more than almost anything else, usually by younger founders, usually with a slightly worried look. And every time, I have the same quiet reaction, which I have learned to keep off my face: the question contains its own mistake. The word balance assumes two weights on a scale, work on one side and life on the other, and a virtuous person keeping them level. I have run companies for over a decade. I have never once felt level, and I have stopped believing the scale exists. What I believe in instead is integration, and I want to be precise about it, because integration is easily misread as a fancy word for never stopping. It is not. It is something more honest and more sustainable than balance ever was. ## Why the scale is the wrong picture A scale implies that work and life are opponents, and that every hour given to one is stolen from the other. For some jobs, maybe that is true. For building a company you care about, it is a miserable and inaccurate way to live, because the best parts of my work do not feel like a withdrawal from my life. They feel like part of it. The problem is not that work takes from life. The problem is pretending the two are cleanly separable in the first place. When I treated them as a scale, I lost both. I was at work feeling guilty about home, and at home feeling guilty about work, and the guilt was the only thing perfectly balanced. ## What integration actually means Integration means I stop trying to keep the two halves apart and instead arrange a single life with a rhythm that bends toward whatever needs me most that week. Some weeks are heavily work. Some weeks, deliberately, are not. A great idea is allowed to arrive on a Sunday walk, and an ordinary Tuesday afternoon is allowed to belong to my family without an apology to anyone. Balance asks you to keep two weights level. Integration lets you admit that some weeks one of them simply matters more, and that is fine. The thing that makes this work, rather than just a permission slip to overwork, is intention. I am not blending the two by accident or by being unable to switch off. I am choosing, week to week, where the weight goes, and then being fully present wherever I put it. Half-present everywhere is the actual enemy, and it is what a rigid idea of balance quietly produces. ## The part that keeps it honest Integration only works if you are ruthless about recovery, otherwise it really does collapse into a permanent low-grade grind. So I am strict about protecting genuine downtime and about defending the hours where the real work happens. The rhythm has to include real rest or it is not a rhythm. It is just a slope. When younger founders ask me how to balance it all, I have started telling them not to. Stop trying to hold the scale level. Build one life with a tempo you can sustain, decide each week what it owes the most to, and be all the way there when you arrive. It is less tidy than balance, and far more livable. I have integrated my work and the rest of my life, imperfectly and happily, for twenty years. I would not survive a week of trying to keep them balanced. The scale was always a trap. The rhythm is the thing. NJ Nikhil Jathar “Balance assumes two equal weights. I have never once felt equal to my weeks.” ---