"52 Repos, Zero Domain Knowledge,
and 6 Weeks to Deliver"

Picture this: Three consultants walk into a client site. None of us know the domain. There are 52 repositories we've never seen. The architecture? "It is complicated." The documentation? "Talk to Sarah, but she is on leave for two weeks." We had six weeks to not just understand this mess, but actually ship something valuable.

We decided to treat AI as our fourth team member — not to replace thinking, but to handle the grunt work so we could think better. This is the story of what happened when we stopped asking "Can AI do this?" and started asking "How do we work with AI to do this?"

You will hear about the mapping technique that turned 52 repos into a system we could actually reason about. The requirements document our product owner called "the best PRD I've ever seen" (her words, not ours). The stupid mistakes we made. The smart things we stumbled into. And the moment we realized we would accidentally built a repeatable way to tackle any complex system.

This is not a pitch for tools. It's an honest account of an experiment that worked better than it had any right to, complete with the templates and techniques we are still using six months later.

What You'll Actually Take Away

• A framework that works when you're drowning in complexity: The three-phase discovery → definition → delivery pattern we used to go from "I have no idea what is happening" to shipping production code in six weeks. With specific prompts and structures, not vague theory.
• How to onboard people (including yourself) at speed: We got team members contributing meaningfully in 3 days instead of 3 weeks. The journey mapping and knowledge capture techniques that made it possible, and why they work even without AI.
• Proven prompting frameworks that changed how we write requirements: Battle-tested prompting strategies using which our PRDs now consistently catch edge cases we used to miss. You'll see it in action.
• Templates you can steal immediately: Repository analysis prompts, PRD structures, and architecture documentation formats. We'll share the actual artifacts, not sanitized examples.
• Where we messed up (so you do not have to): The code review that almost went horribly wrong. The diagram that confused everyone. The time we trusted AI too much. This stuff matters more than the successes.