Field notes from building ProductIntel.
A series on the intersection of product management, AI engineering, and enterprise software, written by someone living through the transition in real time.
These are early previews. I'd rather get the ideas out while the conversation is happening than wait for perfectly polished prose.
- No. 1
Why Claude Code Needs More Than Just Your Words: A Case Study in Story Refinement
A 3-sentence story description turned into a 400-line spec after 30 seconds of AI enrichment. The difference between the two is the difference between an agent that builds what you meant and one that builds what it assumed. Real examples from six months of building ProductIntel.
The evidence·12 min - No. 2
Building an AI Product Manager Skillset, In Public
What 8 weeks of intense AI platform development taught me about the role that doesn't exist yet. Four phases of learning, from the confidence trap to the observability shift, and five things I'd tell myself at Week 0.
The journey·10 min - No. 3
The Pre-Product Discovery Problem
Why the first 5 minutes of an AI project can determine the next 6 months. A tiered discovery framework that surfaces the decisions that are expensive to change later, before the first line of code is written.
The practical tool·10 min
- No. 4
The Product Manager Role Is Splitting in Two. Which Side Are You On?
The conversation about how AI is changing product management is stuck on the easy part. A four-phase framework for the discipline that's actually emerging, and why most PMs will get stuck before they reach it.
The thesis·Coming soon - No. 5
Seven Skills I Believe Will Define the Next Product Manager
If the PM role is really splitting in two, what does the new discipline actually look like? Seven competency domains, concrete enough that a company could hire against them and an individual could develop toward them.
The framework·12 min - No. 6
Your AI Agent Needs an Org Chart, Not Just a Framework
The harness conversation is solving the runtime problem, not the organizational one. A CEO Agent model, borrowed directly from how scaled human organizations actually work, and why the next evolution isn't a better harness.
The contrarian take·Coming soon - No. 7
What 25 Years of Enterprise Product Work Taught Me About AI Agents
Nine months into building multi-agent systems from scratch, the hard part wasn't the new technology. It was realizing how much of the enterprise playbook transferred: handoffs, scoped context, specification rigor, cost discipline, and trust-building.
The bridge piece·Coming soon