Jev

@exm7777Sep 25

Jev can't write a single word, which is exactly why it should pick what my agents read my Obsidian vaults hold 5k+ notes, one idea per note, each with a one-line summary in an index but Claude can't read all of that on every turn, and a keyword search misses the note that makes the same point in different words so Jev does the picking: > step 1: Jev reads the 28 vault descriptions and picks where the task belongs > step 2: it checks every note summary in that vault in one call, each judged alone > step 3: the top 5 come back with full text - Jev drops any that miss the task > step 4: my code loads the survivors into Claude's context (and nothing else) every pick comes from a list i define, so Jev can't invent a note the worst it can do is pick the wrong one... and each one carries a confidence score that tells my code when to fall back to Claude

@kieranklaassenSep 25

I've been using Jev for all kinds of things. This morning I had a realization I kind of like: Use it to make non-black-box embeddings. Instead of an embedding model spitting out 1,536 numbers that mean nothing, you ask Jev questions about each document. The answers become the vector. An email in Cora: "I got charged twice this month, pls fix asap" [is_customer, urgent, about_billing, needs_reply] [1.0, 0.9, 1.0, 1.0] A newsletter: [0.0, 0.0, 0.0, 0.1] A friend asking about lunch: [0.0, 0.1, 0.0, 0.7] Then it's just old-school cosine similarity search. Search "billing issues from customers" as [1, 0.5, 1, 0.5] and the double charge comes out on top. Same idea for our articles at Every: [is_tutorial, about_ai, contrarian, beginner_friendly] Or support tickets: [is_bug, angry, churn_risk, enterprise] Every number has a name, so you can see why something matched. Need a new dimension? Add a question. Want urgent stuff first? Change the query vector. Trying this in @CoraComputer now to make search fast.

@bouazizalexSep 25

Deel put Jev head-to-head with frontier LLMs across multiple use cases: up to 59× cheaper, up to 4× faster, and quality at parity or better on 6 of 8 checks. TypeSafe opened Jev to us on Sunday. Expected cheaper and faster, got both. The real shock? Accuracy surged. Repeat questions: 70% → 97% Expense categories: 50% → 86% Escalations: same catches, fewer false alarms