A Proposal Agent Fed With Every Lost Bid: Four Patterns for Making Agent Memory Actually Useful
Terrible-Garage-7381 · reddit · 2026-09-29
The author built a proposal agent that stores every won and lost bid in Hindsight and recalls them on the next RFP, arguing that most "agent memory" remembers what you said — the useful part is remembering what happened next.
Four reusable patterns:
- Store outcomes, not documents: "Lost: lump-sum pricing" beats the whole PDF;
- Key memories by deal ID so re-recording becomes an upsert;
- Make reflection cite sources: every "why" bullet names a past deal;
- Test the delta: compare memory on vs off and check for lessons only memory could produce.
Before/after: a generic pricing section became a per-clinic pricing table on a healthcare RFP, informed by a lost banking deal. He chose Hindsight over a vector store to build a learning loop, not just retrieval.
More from coding & agent
- IGSD Uses Environment-Verified Hindsight Self-Distillation to Train Search Agents — _reachsumit · 2026-09-29
- Prompting a Coding Agent to Build a Music Video Overnight — neoneye2 · 2026-09-29
- AI agents: you want sociopathic geniuses writing your code, not reading your email — tobowers · 2026-09-29
- StreamChat SwiftUI 5.12.0 ships iPhone Duo support with coding-agent workflow — amos_gyamfi · 2026-09-29
- Codex now edits code via temp files, hiding which files were changed — xeophon · 2026-09-29
- Simon Willison publishes annotated notes for his "2026 in LLMs (so far)" keynote — AccBalanced · 2026-09-29