Async

Final system brief and production-readiness checklist

Write a one-page brief a stranger can understand:

  • Problem or product: what you are solving (fun projects are fine; say what it does and why it is interesting).
  • Architecture: agents, memory, tools, APIs, and how the pieces talk.
  • Stack and tools: FastAPI, LangGraph/ADK, RAG store, memory store, eval harness, host.
  • Evals: what TRACE proved, and at least one fix you shipped from failures.
  • Memory: what you keep, when you write, where it lives, how you retrieve, when you forget.

Plan a 5 to 10 minute Demo Day walkthrough. Aim for about 5 minutes of live demo inside that slot. Keep the story aligned: problem → architecture → stack → live demo.

Critical

Record a backup demo

Live demos break. Record a short screen capture of the same flow. Bring it to Demo Day as insurance.

Production-readiness checklist

  • URL loads for a stranger (incognito) without your laptop VPN tricks
  • Core task works end to end
  • Memory persists across a fresh session
  • Eval suite still passes (or you can show the latest score + known gaps)
  • At least one fix from TRACE is shipped
  • README or brief covers problem, architecture, stack, and demos
  • Backup recording exported and ready