Harsha Cheruku · @fullstackpm
The PM
who ships.
PM Reading Stack
What I'm reading, listening to, and thinking about this month · View archive
The Product Operating Model
Still one of the cleanest references for moving from feature teams to empowered product teams. Useful calibration before org design debates.
How to Build Better Products Through Continuous Discovery
Teresa Torres lays out practical discovery cadence. Great reset when roadmap delivery starts swallowing customer learning.
Good Strategy / Bad Strategy (Kernel Refresher)
Revisiting diagnosis → guiding policy → coherent actions. Helps keep PM docs from turning into goal lists without leverage.
Satya Nadella on AI Platform Shifts
Useful lens on platform transitions, ecosystem moats, and why distribution + workflow integration matters more than raw model demos.
Lenny Rachitsky: Product Sense and Career Compounding
Bookmarking the channel for interviews that blend tactical PM playbooks with operator-level thinking.
INSPIRED (2nd Edition)
Re-reading selected chapters with an AI-era lens: what changes, what fundamentals remain brutally unchanged.
Fresh from the feed
Live Tools
Built & shipped, not just plannedPM Multiverse
Same problem. 5 PM brains. Where do Cagan, Torres, Doshi, Lenny, and the Exec AI Monster disagree — and which one thinks like you?
10 problems · 5 personas · Live community votes →
🎯 Interview Coach
Practice PM interviews. Get AI feedback on your frameworks, structure, and thinking. Build confidence before the real thing.
Start practicing →
📋 SDE Prep
9-week job search intensive. 200+ daily tasks, structured to get you from prep mode to signed offer.
See the plan →
Latest Writing
All articles →When P Goes Wrong: Case Studies of Catastrophic Probabilistic Failures
P 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.
2031 — Part 4: The Number
Kai 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.
The Economics of P: Who Gets Paid What When D Gets Cheaper?
When 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.
Stay in the loop
PM insights, shipped projects, and lessons learned — straight to your inbox.
@fullstackpm
PM instinct.
Ship mindset.
Built by a PM who codes. 16 years of product experience, now shipping fun projects in public.