Unblocked's relational context engine cuts agent token costs by ~50%
AI Engineer · youtube · 2026-10-11
Peter Werry, founding engineer at Unblocked, argues coding agents need more than vector search: a relational context engine can answer structural questions ("which PRs did I merge last week?") using PRs, Slack threads, and incident history.
- Real Claude sessions with this context showed 50% cost reduction, and context helped trace a regression to a model switch
- Live build of Unblocked's open-source document query engine: automatic schema discovery, identity resolution across systems, query validation blocking dangerous operators, error-feedback retries, and full-text search
- Ships with an open-source context engine simulator to test whether pre-gathered context helps your agents
More from coding & agent
- Free Photoshop clone surfaces online with control ports built for AI agents — aakashgupta · 2026-10-11
- Use Claude and Codex to build a 'design lab' that centralizes design iteration across projects — RileyRalmuto · 2026-10-11
- gemini-cli PR fixes ACP session/load responding before history replay completes — Nodge · 2026-10-11
- Three agents burned $47.35 overnight: uncapped loop instructions are a billing trap — aakashgupta · 2026-10-11
- Stop Calling Yourself 'Just a Vibe Coder': Learn Enough That Agents Can't Fool You — brandon_galang · 2026-10-11
- Agents Don't Read Your Policy Docs: Gravitee demos gateway-level AI governance — AI Engineer · 2026-10-11