Writing by Tarun Upaday on company-building, operational software, applied AI and travel.https://tarun.me/en-ushi@tarun.me (Tarun Upaday)hi@tarun.me (Tarun Upaday)https://tarun.me/posts/an-ai-task-is-not-done-until-the-world-has-changed/https://tarun.me/posts/an-ai-task-is-not-done-until-the-world-has-changed/I moved my working life onto AI agents and logged every failure. Each one traced to the same cause: the to-do item is the wrong primitive for a machine.Sun, 16 Aug 2026 00:00:00 GMTai-agentsverificationagent-operationstraceguardhi@tarun.me (Tarun Upaday)https://tarun.me/posts/pillwatch-when-99-percent-is-not-enough/https://tarun.me/posts/pillwatch-when-99-percent-is-not-enough/Four high schoolers, a $60 Raspberry Pi, and the compounding math that turns 99 percent per cell into an error every fifth box.Sun, 16 Aug 2026 00:00:00 GMTpillwatchcomputer-visionedge-aireliabilityteachinghi@tarun.me (Tarun Upaday)https://tarun.me/posts/pillwatch-data-small-honest-datasets/https://tarun.me/posts/pillwatch-data-small-honest-datasets/The PillWatch team labels its own data, splits it by photograph, and hunts its rare cases deliberately. That discipline taught more than model tuning did.Sun, 16 Aug 2026 00:00:00 GMTpillwatchdatasetslabelingopen-datahi@tarun.me (Tarun Upaday)https://tarun.me/posts/the-llm-call-you-can-delete/https://tarun.me/posts/the-llm-call-you-can-delete/631 replayed decisions, six written rules, zero disagreements. The model I was paying to route my email lost to the policy file it was reading.Sun, 16 Aug 2026 00:00:00 GMTai-agentsevaluationdeterministic-offloadagent-operationshi@tarun.me (Tarun Upaday)https://tarun.me/posts/tim-cooks-legacy/https://tarun.me/posts/tim-cooks-legacy/Most CEOs leave too late. Cook's real legacy is not the products. It is the restraint, and the impeccable timing on the way out.Thu, 23 Apr 2026 00:00:00 GMTappletim-cookleadershiptechhi@tarun.me (Tarun Upaday)https://tarun.me/posts/pivot-gates/https://tarun.me/posts/pivot-gates/How founders can distinguish a slow strategy from a broken premise before sunk cost makes the decision for them.Tue, 31 Mar 2026 00:00:00 GMTroutespringpivotdecision-makingai-pivothi@tarun.me (Tarun Upaday)https://tarun.me/posts/travel-routing-optimization/https://tarun.me/posts/travel-routing-optimization/The hard part of flight routing was never finding the cheapest itinerary. It's that 'cheapest' and 'best' are different questions, and one of them is a hard rule while the other is a matter of taste.Mon, 19 Jan 2026 00:00:00 GMTtravelroutingapplied-aiproducthi@tarun.me (Tarun Upaday)https://tarun.me/posts/technical-system-troubleshooting/https://tarun.me/posts/technical-system-troubleshooting/Most production failures in integrated systems aren't yours or theirs. They're the handoff between the two, and that's exactly where nobody looks first.Mon, 08 Dec 2025 00:00:00 GMToperating-notestroubleshootingenterprise-softwareintegrationshi@tarun.me (Tarun Upaday)https://tarun.me/posts/build-versus-buy-decisions/https://tarun.me/posts/build-versus-buy-decisions/The standard framing, comparing the cost of building to the cost of buying, misses the question that actually decides it: does the vendor's product expose the interface you need, or just the outcome?Mon, 17 Nov 2025 00:00:00 GMToperating-notesbuild-vs-buyproduct-strategyapplied-aihi@tarun.me (Tarun Upaday)https://tarun.me/posts/operating-notes/what-exits-teach-you/https://tarun.me/posts/operating-notes/what-exits-teach-you/An exit gives a founder credibility, not certainty. What building and selling companies actually teaches, and why a previous exit can quietly become a liability.Mon, 03 Nov 2025 00:00:00 GMToperating-notesexitscompany-buildinghi@tarun.me (Tarun Upaday)https://tarun.me/posts/operating-notes/how-buyers-evaluate-operational-risk/https://tarun.me/posts/operating-notes/how-buyers-evaluate-operational-risk/Founders assume buyers are evaluating the product. They're also evaluating what happens when it fails. Why an inferior incumbent keeps winning, and how to sell controlled change instead of pure improvement.Mon, 27 Oct 2025 00:00:00 GMToperating-notesenterprise-salesenterprise-softwarehi@tarun.me (Tarun Upaday)https://tarun.me/posts/operating-notes/services-wedge-or-trap/https://tarun.me/posts/operating-notes/services-wedge-or-trap/Services can be the only credible way into a complex industry, or the thing that caps your margins forever. The distinction isn't whether humans are involved. It's whether their effort compounds.Mon, 20 Oct 2025 00:00:00 GMToperating-notesservicesenterprise-softwarehi@tarun.me (Tarun Upaday)https://tarun.me/posts/operating-notes/exception-handling/https://tarun.me/posts/operating-notes/exception-handling/Enterprise software is designed around states. Real operations happen between them. Why systems fail at the moment users most need help, and what good exception handling requires.Mon, 13 Oct 2025 00:00:00 GMToperating-notesenterprise-softwareexception-handlinghi@tarun.me (Tarun Upaday)https://tarun.me/posts/operating-notes/manual-minutes-per-transaction/https://tarun.me/posts/operating-notes/manual-minutes-per-transaction/"AI adoption" is one of the least useful metrics in software. Manual minutes per transaction is hard to fake, captures hidden labor, and separates adopting AI from actually improving operations.Mon, 06 Oct 2025 00:00:00 GMToperating-notesapplied-aimetricshi@tarun.me (Tarun Upaday)https://tarun.me/posts/operating-notes/is-this-workflow-automatable/https://tarun.me/posts/operating-notes/is-this-workflow-automatable/Every workflow looks automatable from a distance. Up close it's a sequence of judgments, not clicks. Five questions to ask before you automate anything, and why the process that works 80 percent of the time is the most dangerous kind.Mon, 29 Sep 2025 00:00:00 GMToperating-notesautomationapplied-aihi@tarun.me (Tarun Upaday)https://tarun.me/posts/operating-notes/one-customer-dominates/https://tarun.me/posts/operating-notes/one-customer-dominates/Concentration isn't the real risk. Failing to convert a dominant customer into reusable product is. The metric that matters isn't percentage of revenue. It's percentage of the business that could transfer to another customer.Mon, 22 Sep 2025 00:00:00 GMToperating-notesenterprise-softwarecustomer-concentrationhi@tarun.me (Tarun Upaday)https://tarun.me/posts/operating-notes/company-building-under-real-constraints/https://tarun.me/posts/operating-notes/company-building-under-real-constraints/A seven-part series on building operational software companies: customer concentration, automation, exception handling, services, enterprise risk, and what exits do and don't teach. Written from the operator's chair, not the slide deck.Mon, 15 Sep 2025 00:00:00 GMToperating-notesenterprise-softwarecompany-buildinghi@tarun.me (Tarun Upaday)https://tarun.me/posts/archive/2009-11-17-2-things-every-good-startup-needs/https://tarun.me/posts/archive/2009-11-17-2-things-every-good-startup-needs/Robert Scoble has a great article today on his take on mistakes startups make .Sat, 21 Nov 2009 00:00:00 GMTstartuphi@tarun.me (Tarun Upaday)https://tarun.me/posts/archive/2009-11-18-when-good-people-resign-to-join-you/https://tarun.me/posts/archive/2009-11-18-when-good-people-resign-to-join-you/You make an offer to an A+ player at a big company, they accept — and then their employer rolls out the red carpet to keep them. Mark Suster calls it the counter-offer, and it's an all too familiar phenomenon worth planning for before you extend the offer.Wed, 18 Nov 2009 00:00:00 GMTwritinghiringhi@tarun.me (Tarun Upaday)https://tarun.me/posts/archive/2009-09-16-why-web-designers-are-over-rated/https://tarun.me/posts/archive/2009-09-16-why-web-designers-are-over-rated/Wired did a competition inviting designers to improve craigslist .Mon, 28 Sep 2009 00:00:00 GMTuncategorizedhi@tarun.me (Tarun Upaday)