@ADLXBT @pedronauck @typesafeai Jev is insanely fast though, usually around 200
@ADLXBT @pedronauck @typesafeai Jev is insanely fast though, usually around 200 ms, which is pretty trivial for coding agents.
Crawled from X, GitHub, blogs, and docs. Jev classifies, de-duplicates, and ranks. Every row links back to the original source.
Fresh crawls from X, GitHub, and the web. Newest first.
@ADLXBT @pedronauck @typesafeai Jev is insanely fast though, usually around 200 ms, which is pretty trivial for coding agents.
@BhosalePratim @typesafeai @GradiumAI acting on a prefix only works if the wrong tool is cheap to undo. for voice, do you wait for a Jev confidence cutoff, or fire as soon as one t
been playing with Jev from @typesafeai and built unsaidbrief the idea: we give AI builders vague prompts, then spend hours fixing decisions we never actually made paste your brief
@instantricecook @typesafeai My Jev app just classified this post as complete bullshit.
@moritzkremb Has anyone tried using Jev inside a speculative decoding pipeline — e.g. generate many candidate continuations cheaply, use Jev to rank them in parallel, and only have
@mandel59 match はパターンマッチで、noul, choice, score は jev 側のモデル抽象です。 一応アップロードしてますが、真面目に作ってないです https://t.co/MnfGTu5Zpi

You can have Jev at home now!
Introducing Jev Reviewer, a tool designed for systematic reviewers to quickly find and extract the information from research articles.
made this in under an hour to play around with jev.
@ZacGawn Try this memory system, better than obsidian. Obsidian was never built for ai. Engraphis works great with regular ai and Jev https://t.co/dkha9PnUTp
https://t.co/3FN9fy008i openevals with jev cheaper and faster
jev でメモリコンパクションを実装する例。ログが今の作業と関係ある/ないを分類してフィルタするっぽい。 https://t.co/fe7idXMu95
更新しました。 -New- ハエの脳 →キノコ体回路に限定 Jev →未来の限定的な中間supervisor 実コネクトームからキノコ体の一部を抽出して使う限定的な学習実験。 #個人開発 https://t.co/9oBkYWzIFA

TypeSafe raised $40M for Jev: typed decisions in one forward pass, calibrated. Closed weights, waitlist. I'm building the open version on one RTX 3090: Qwen3.8-27B, Apache-2.0, ev
@sydneyrunkle @ktech9999 Jev + sealed trace ledgers feels like a natural pair — we're stamping agent tool calls into tamper-evident NEEDLE DROP chains so a harness run leaves check
@typesafeai https://t.co/dOsKPoWzJ3 Point it at an MCP server and it automatically probes the tool signatures, then runs a series of typed Jev calls to pick the right function and
@hov8a @typesafeai Yep this is one I’m building: https://t.co/ZBWc616ww6 JEV does the fast semantic judgments, deterministic code handles ALLOW / REVIEW / HOLD.

Introducing Bespoke Nimble: an open data, open model, open recipe for an open Jev.
@rodrigovmedia @antonioleivag Yo me armé esta tool para que todo el plan viva en el disco, con soporte multiagente, resume, switch, etc, debería agregarle algún soporte para Jev, p
https://t.co/CxITxpE1Ss `jev` branch