The One Rule for Trustworthy Quoting Agents: Never Let the LLM Touch the Numbers
Futdragun · reddit · 2026-10-01
A developer building quoting agents for custom-work businesses shares his architecture: LLMs only extract (temperature 0, strict JSON), missing fields stay null and become customer questions, pricing runs on a deterministic code engine, price regression tests must match to the cent, and a human approves everything. His key lesson: the hard part isn't the AI but the domain vocabulary customers leave out. The pattern—extract, flag, ask, price with code, human approves—applies to any quoting business.
More from coding & agent
- All the Copyright Easter Eggs in That AI Music Video Came From Claude — technollama · 2026-10-01
- ChatGPT Wrote Lyrics, Suno Made Music, Claude Coded a 5,200-Frame Music Video — technollama · 2026-10-01
- New tool makes benchmarking across factory configs trivial, not just models — vikvang1 · 2026-10-01
- Magnitude inference engine hits #1 on HN, claims up to 2x faster local open-model runs than llama.cpp — nickbaumann_ · 2026-10-01
- marimo-lens lets you point at charts and steer your coding agent directly — S_Conradi · 2026-10-01
- Non-coder builds MCP-only medieval trade game played entirely by AI agents — Wainfare · 2026-10-01