There's a lot of excitement right now about product managers becoming AI builders. I've been living that reality at Origin, shipping features for patient outcomes tracking that go well beyond what I'd typically advise a PM to take on. And it's been working — but I think it's worth being honest about why.
It's not just because AI makes building easier. It's because of three things I brought to the table before I wrote a single line of code.
I deeply understand the problem. I've been working hand in hand with our clinicians for five years. I know the workflows, the pain points, and the nuances that don't show up in any spec. My clinical thought partner Liz Miracle, MSPT, WCS has been invaluable here — this kind of domain depth isn't optional, it's table stakes.
I deeply understand the data. I've built many of the underlying tables in BigQuery myself. I understand not just what the data means, but the API calls required to move it between systems, because I built those foundations. That's not typical for a PM — and I think it's the real differentiator in why this has worked.
I know where my lane ends. Security, production-level integration, infrastructure decisions — that's Nick Kavassalis's territory as CTO, full stop. The goal was never to do it all. It was to go as far as I responsibly could, and hand off with enough context to make the collaboration count.
AI has genuinely expanded what's possible for product leaders who want to build. But the ceiling on that isn't the tools, it's the depth you bring before you open Claude Code.
What are you seeing in your orgs? Are product leaders picking up the builder mantle? What's working?