AI PM Readiness Diagnostic

Most stalled AI products aren't stalled on model quality. They're stalled on the operating layer underneath — and after 15 years running product in enterprise SaaS, regulated financial data, and healthcare, I see the same three failure modes over and over:

Context debt. The knowledge required to make good decisions about this product lives in six people's heads, four Slack threads, and a deck from March. Every new stakeholder resets the conversation. Every handoff loses fidelity. Your team is re-deciding things it already decided.

Alignment tax. Legal, security, and compliance are treated as a gate at the end instead of a constraint at the start. So you build, then discover the thing you built can't ship in your environment, then rebuild. The cost isn't the rework — it's that your team stops proposing ambitious things.

Strategic drift. The AI initiative was funded against one business outcome and is now being measured against a different one. Nobody said this out loud. Everyone can feel it.

This session diagnoses which of the three is actually costing you, and what to do about it in the next 30 days.

This is for you if: you own an AI or AI-adjacent product inside a company with real constraints — enterprise scale, regulated industry, security review, procurement. You have a live initiative that is moving slower than it should and you can't fully articulate why.

This is not for you if: you're pre-idea and want help choosing what to build, you're looking for technical architecture review, or you want a career conversation. Good things — just not this session.

No slides. No framework lecture. You bring one real initiative and we work on it.

Minutes 0–15 — The actual state. You walk me through where the initiative stands: what's built, who's involved, what decision is currently pending, and what you've been told the blocker is. I ask questions that are often uncomfortable. This part is deliberately unstructured.

Minutes 15–45 — Diagnosis. We locate the real constraint. Usually it's not the one named in the status update. I'll map what you've described against the three failure modes above and show you which one is dominant, which is secondary, and which is a symptom rather than a cause. Where I've seen your exact situation before — at Oracle scaling a platform past 15M users, or building 0→1 in regulated financial services at S&P Global — I'll tell you what worked and what didn't.

Minutes 45–70 — The 30-day path. We build the sequence: what you change first, what you deliberately leave broken, who you need in the room, and what the conversation with your leadership needs to sound like. Sequencing matters more than the list. Most teams try to fix all three at once and fix none.

Minutes 70–75 — What I'd watch for. The two or three things most likely to knock this off course, so you see them coming.

The session is direct. If your plan has a hole in it, I'll tell you. That's what you're paying for.

Within 48 hours you receive:

• A one-page written diagnostic naming your dominant constraint and the evidence for it — in language you can forward to your manager or bring to a leadership review without translating it first

• A prioritized 30-day action list, sequenced, with the first move called out explicitly

• The specific stakeholder conversations that need to happen, and who needs to be in each one

• Any templates or artifacts from my own toolkit that fit your situation

The one-pager is written to be circulated. Several people have used it as the basis for a re-scoping conversation with their leadership.

To prepare, send me in advance (optional but it makes the session materially better):

• One artifact — a PRD, roadmap, architecture note, or even a status deck. Whatever actually represents the initiative.

• Two or three sentences on what you think the problem is.

I'll review it before we meet so we don't spend session time on context.

What this is not: it isn't a sales call for a longer engagement, and it isn't a generic AI strategy overview. It's 75 minutes on your specific problem, with a written output. If you decide afterward that you want ongoing support, we can talk about it — but the diagnostic stands entirely on its own.