AI product management
I define AI products and ship them.
The job is not the spec. It is alignment: stakeholders who trust the plan, engineers who respect it, and users who adopt what ships. I write the PRD, define what success measurably means, and carry the product through adoption.
What that looks like
Product strategy and roadmaps
Where AI actually pays off in your workflow, sequenced, with the dependencies stated plainly.
Specs engineers respect
PRDs with acceptance criteria, written by someone who also writes the code.
Metrics and evaluation design
A definition of working you can defend to a budget holder, and the pipeline that measures it.
Zero-to-one platform builds
From first conversation to production, with the operational runbook included.
Case studies
Editorial AI platforms
Two platforms inside a large news publisher: a self-service automation layer and a reusable embed system, across 9 markets.
Read the case study →AI visibility measurement platform
Built solo for a consulting client: ~1.35M AI responses collected a month across six engines.
Read the case study →
Journalists
Journalists don't need AI to do their writing; but they can use AI tools to deepen their research, reveal thoughtful angles, and find valuable stories in complex data. The AI Upgrade for Journalists course introduces the best of these tools, and teaches the best ways to embed them into the reporting process.
Have an AI product that needs to exist, or one that exists and is not being used?
Email peter@peterbittner.com