v0.4 · updated August 2026 · changelog
Two practical playbooks for startup and scale-up product teams. Every step is backed by a recognised framework. So when someone asks "says who?", the answer is on the page.
AI is in the thing you ship. Problem to live to well-operated.
AI is in how you build. Ship faster without losing the plot.
<aside> 📦
Everything lives on GitHub: full phases, exit criteria and fill-in templates. github.com/EmilieBuilds/AIProductPlaybook
</aside>
| Source | What it is | What the playbooks take from it |
|---|---|---|
| NIST AI Risk Management Framework | The US government's framework for managing AI risk. The most cited AI risk reference in the world, and free. | The risk backbone: Govern, Map, Measure, Manage. |
| ISO/IEC 42001 | The international standard for managing AI responsibly. Companies can be certified against it. Anthropic is. | Clear accountability, impact assessments, good documentation habits. |
| DORA / Accelerate | The biggest research program on what makes software teams perform. | Ship small and often. Speed and stability rise together, not against each other. |
| Anthropic's published guidance | Practical guides from a frontier AI lab on building with AI. The principles work with any model provider. | Start simple. Test before you trust. Treat prompts like code. |
| Regulation (overlay) | The EU AI Act, plus TGA (Australia) and FDA (US) rules for medical software. | Checkpoints marked ⚕️ for products in regulated areas like health. |
Maintained by EmilieBuilds. Licensed under CC BY 4.0: share and adapt with credit. The playbooks map to the named frameworks and do not reproduce their text. Not legal or regulatory advice.