The AI Product Playbook

What's Inside

1. 1 AI as Tool vs. AI as Product

   A decision framework for when AI should be something you build with versus something you sell. Includes the two-axis test and a decision tree.

2. 2 The Value Proposition Shift

   What changes about your pitch when AI is the core. How to position against non-AI alternatives and other AI products.

3. 3 Structured Outputs Beat Magic

   Why deterministic AI products win over impressive demos. How to design AI features that feel reliable enough to depend on.

4. 4 The Eval Problem

   How to know if your AI product is actually good. Minimum viable evals, scoring rubrics, and when to use LLM-as-judge.

5. 5 Choosing Your AI Architecture

   API vs. open-source vs. fine-tuned. When you need RAG. How to estimate your AI costs per user per month.

6. 6 Pricing AI Products

   The margin math is different. Usage-based, tiered, and outcome-based models — with real cost calculations.

7. 7 The Wrapper Trap

   When you're one API change away from death. Four defenses against wrapper risk, and how to audit your own exposure.

8. 8 Building Trust with Non-Technical Users

   Your users don't care about your model. How to build trust incrementally — from suggestions to decisions.

9. 9 Case Teardowns

   Real AI products analyzed through every framework in the playbook. What worked, what didn't, and why.

10. 10 Your 2-Week AI Product Sprint

    A day-by-day action plan to validate your AI angle, build one feature with structured outputs, and ship an MVP.

Who This Is For