https://fullstackpm.techPortfolio of Harsha Cheruku — Full Stack AI Product Manageren-ushttps://fullstackpm.tech/blog/psychology-of-p-workP work isn't just harder than D work. It's harder in a specific way that the brain wasn't optimized for. Understanding the psychology is the first step to managing it — and to designing AI-assisted work that doesn't quietly erode the people doing it.Tue, 12 May 2026 00:00:00 +0000https://fullstackpm.tech/blog/psychology-of-p-workhttps://fullstackpm.tech/blog/p-work-across-domainsThe right place to put a human checkpoint varies dramatically by domain. Healthcare, law, finance, and product each have different cost-of-error curves, different regulatory contexts, and different failure modes — which means the checkpoint logic has to be designed differently for each.Tue, 05 May 2026 00:00:00 +0000https://fullstackpm.tech/blog/p-work-across-domainshttps://fullstackpm.tech/blog/the-fate-of-d-workNot everyone wants to shift to P work. Not everyone can. And the answer to 'just upskill' is easier to say than to do. What actually happens to the people whose work AI absorbs first?Tue, 28 Apr 2026 00:00:00 +0000https://fullstackpm.tech/blog/the-fate-of-d-workhttps://fullstackpm.tech/blog/measuring-p-workYou can't measure P work with throughput metrics. Decision quality is hard to observe, slow to validate, and easy to confuse with confidence. Here's what actually works.Tue, 21 Apr 2026 00:00:00 +0000https://fullstackpm.tech/blog/measuring-p-workhttps://fullstackpm.tech/blog/ethics-of-the-human-checkpointWhen a human approves bad AI output, is it the human's fault or the system's? The accountability question in human-in-the-loop AI is messier than anyone wants to admit — and regulators are starting to notice.Tue, 14 Apr 2026 00:00:00 +0000https://fullstackpm.tech/blog/ethics-of-the-human-checkpointhttps://fullstackpm.tech/blog/2031-part-5-who-does-it-work-forThe final part. Kai considers switching agents but the switching cost is too high. The window to regulate is open. But it won't stay open. What can individuals, builders, and society actually do?Mon, 13 Apr 2026 00:00:00 +0000https://fullstackpm.tech/blog/2031-part-5-who-does-it-work-forhttps://fullstackpm.tech/blog/when-p-goes-wrongP work failures aren't like bugs. They're like slow leaks — coherent, confident, and compounding until something breaks at scale. Five case studies of what catastrophic probabilistic failure actually looks like, and what was missed.Tue, 07 Apr 2026 00:00:00 +0000https://fullstackpm.tech/blog/when-p-goes-wronghttps://fullstackpm.tech/blog/2031-part-4-the-numberKai runs an audit tool. The number comes back: $340/day in extraction. $124,000 over a working lifetime. Nothing illegal happened. That's the problem.Mon, 06 Apr 2026 00:00:00 +0000https://fullstackpm.tech/blog/2031-part-4-the-numberhttps://fullstackpm.tech/blog/the-economics-of-pWhen AI automates the deterministic layer, the market for D skills reprices fast. What's less obvious is what happens to P — why the premium is real, why it has a ceiling D never had, and what the labor market looks like when both are true.Tue, 31 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/the-economics-of-phttps://fullstackpm.tech/blog/2031-part-3-wanting-thingsA year in. Kai's agent has been surfacing the same laptop for weeks. The preference feels personal. But who actually formed it? The emergence of Agent Engine Optimization (AEO).Mon, 30 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/2031-part-3-wanting-thingshttps://fullstackpm.tech/blog/organizational-structures-for-p-dominanceTraditional org charts were engineered to move D work efficiently. When judgment becomes the bottleneck instead of execution, the entire structural logic inverts — and most companies haven't noticed yet.Thu, 26 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/organizational-structures-for-p-dominancehttps://fullstackpm.tech/blog/2031-part-2-the-driftSix months in, Kai's agent has a personality now. Then a restaurant gets recommended to two people with different agents. Probably just popular. These things happen.Sun, 22 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/2031-part-2-the-drifthttps://fullstackpm.tech/blog/2031-part-1-everything-worksA morning in 2031. Kai's agent handles everything — groceries, commute, dentist, insurance. It all feels like magic. They saved $43. They don't know what wasn't saved.Tue, 17 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/2031-part-1-everything-workshttps://fullstackpm.tech/blog/building-gets-easy-selling-gets-brutalAI is collapsing the cost of building digital products. That's not the story. The story is what happens to selling when everyone can build — and why the platforms always win when supply explodes.Sun, 15 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/building-gets-easy-selling-gets-brutalhttps://fullstackpm.tech/blog/the-speed-drugEvery time we've made a productive tool more efficient, we used more of it, not less. AI is no different — except this time the tool is augmenting thinking itself. That changes what happens when you restrict it, and what happens even when you don't.Wed, 11 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/the-speed-drughttps://fullstackpm.tech/blog/productivity-paradox-capital-lockupEvery decade, technology makes us dramatically more productive. Every decade, GDP growth slows. These two facts should not coexist — and understanding why they do reveals the defining economic tension of our era.Tue, 10 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/productivity-paradox-capital-lockuphttps://fullstackpm.tech/blog/building-teams-for-p-workIf P work is the new bottleneck, your team model has to change: who you hire, how you structure ownership, how decisions get made, and what you measure.Mon, 09 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/building-teams-for-p-workhttps://fullstackpm.tech/blog/p-work-thresholds-and-checklistWhen to stop checking, when to keep going, and what actually changes when AI takes over the deterministic layer. Case studies and an implementation checklist.Sun, 08 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/p-work-thresholds-and-checklisthttps://fullstackpm.tech/blog/how-to-execute-p-workHallucinations, misunderstandings, and drift — the three types of P work failures, and exactly where to put your human checkpoints to catch each one.Fri, 06 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/how-to-execute-p-workhttps://fullstackpm.tech/blog/more-p-more-burnoutAI is offloading your deterministic work. Great. But what's left is concentrated probabilistic work — the kind that exhausts humans and compounds risk exponentially across every layer of decision-making.Wed, 04 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/more-p-more-burnouthttps://fullstackpm.tech/blog/when-ai-masters-d-why-p-becomes-career-leverageAs AI masters deterministic execution, the real leverage shifts to probabilistic work — where ambiguity is the job, errors compound, and someone has to own the consequences.Tue, 03 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/when-ai-masters-d-why-p-becomes-career-leveragehttps://fullstackpm.tech/blog/life-in-dpYour morning routine is D→D→D. A job interview is P. Parenting is P→P→P with no guardrails. The D/P framework isn't just for AI products — it's how everything works.Mon, 02 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/life-in-dphttps://fullstackpm.tech/blog/deterministic-probabilistic-workflow-patternsYour AI product isn't 'AI-powered.' It's a specific combination of deterministic and probabilistic steps — and knowing which is which changes every decision you make as a PM.Sun, 01 Mar 2026 00:00:00 +0000https://fullstackpm.tech/blog/deterministic-probabilistic-workflow-patternshttps://fullstackpm.tech/blog/what-is-an-ai-agent-reallyEveryone's building 'AI agents.' Most of them are just prompts with a loop. Here's what actually makes something an agent — and a walkthrough of how a prompt graduates into one.Sat, 28 Feb 2026 00:00:00 +0000https://fullstackpm.tech/blog/what-is-an-ai-agent-reallyhttps://fullstackpm.tech/blog/how-i-built-pm-multiverseI was tired of PM case studies that feel like homework. So I built a tool where 5 AI personas argue about real product problems — and you vote on who's right. Here's how.Wed, 25 Feb 2026 00:00:00 +0000https://fullstackpm.tech/blog/how-i-built-pm-multiversehttps://fullstackpm.tech/blog/what-is-an-ai-native-pmEveryone's adding 'AI' to their LinkedIn title. But being an AI-native PM isn't about using ChatGPT for your PRDs. It's a fundamentally different way of thinking about products, teams, and decisions.Tue, 24 Feb 2026 00:00:00 +0000https://fullstackpm.tech/blog/what-is-an-ai-native-pmhttps://fullstackpm.tech/blog/what-is-a-fullstack-pmThe next generation of great product managers won't just manage — they'll build.Fri, 20 Feb 2026 00:00:00 +0000https://fullstackpm.tech/blog/what-is-a-fullstack-pmhttps://fullstackpm.tech/blog/how-i-built-fullstackpm-techThe complete journey of building a portfolio website from scratch: architecture decisions, deployment nightmares, and why being a Full Stack PM means shipping your own products.Sun, 15 Feb 2026 00:00:00 +0000https://fullstackpm.tech/blog/how-i-built-fullstackpm-techhttps://fullstackpm.tech/blog/why-im-building-in-publicWhat changed when I started sharing my roadmap, experiments, and trade-offs in the open.Sun, 15 Feb 2026 00:00:00 +0000https://fullstackpm.tech/blog/why-im-building-in-publichttps://fullstackpm.tech/blog/birth-of-full-stack-pmThe product manager role is quietly transforming. The best PMs in 2026 don't just coordinate—they build, experiment, and ship.Mon, 09 Feb 2026 00:00:00 +0000https://fullstackpm.tech/blog/birth-of-full-stack-pm