A quiet, daily-updating gallery of what people are doing with Jev — the "System One" decision model from TypeSafe AI that takes a state and typed questions (Choice, Score, Noul) and returns calibrated answers in milliseconds.
Every morning the tracker searches X, Reddit, Hacker News, YouTube, GitHub, TypeSafe's own blog, docs and repos, and the open web. It keeps four kinds of item — concrete use cases, open-source code that builds on Jev, substantive opinions (with the author's verdict), and official updates — drops the link-drops and brochure restatements, and files each card under the topic it is about: routing, extraction, guardrails, agent logic, real-time, tooling, benchmarks, pricing, research. Clicking a card opens the original post or repo.
Categories are yours to shape: rename them, recolour them, reorder them, move a card that landed in the wrong bucket, change its kind, pin the good ones and hide the noise.