Polylane swapped LLMs for decision model Jev in prod, cutting costs 39%
multiply_matrix · x · 2026-09-26
Polylane replaced LLMs with TypeSafe AI's decision model Jev to power its always-on on-call agent, reporting a 39% cost reduction.
- What Jev is: a decision model that generates no text—it answers questions about your data with typed values and calibrated probabilities, at 70–500ms per response with free output tokens, pitched as "smart if-statements" for classify/route/score steps.
- Three question types: Noul (yes/no probability), Choice (pick an option), Score (probability-weighted rubric rating).
- Use case: deciding when its Slack/GitHub agents should proactively respond with meaningful insights.
- The team says for what it's good at, Jev "absolutely mogs" open-source models, and it has now run in production for a week.
More from coding & agent
- Sydney Codex Community Buildathon set for Nov 14 with Cerebras-backed ultrafast compute track — paw_lean · 2026-11-11
- shadcn's improve hits 10k stars: let top models plan, cheap models execute code fixes — shadcn · 2026-11-11
- OpenAI open-sources Codex Security CLI and SDK for AI-driven vulnerability scanning and fixing — openai · 2026-11-11
- Claude Opus 5.5 runs an architecture search and finds AttnRes-Canon, a new top Transformer design — SonglinYang4 · 2026-11-11
- 5 research habits worth knowing: read papers backwards, reproduce first, wire coding agents via MCP — cneuralnetwork · 2026-11-11
- 'Adam can now do CAE': AI agent showcase in engineering simulation draws buzz — burhop · 2026-11-11