jevbooks

← All projects

jev-ultralightspeed

collapseindex/jev-ultralightspeed · Homepage

Batch typed judgements over text queues with TypeSafe's Jev.

Classification & Routing100%Runner-up: Verification & Guardrails

What it is

A Python library that answers one question about each item in a large pile of text — tickets, reviews, logs, model completions — by packing many items into each Jev request. For queues, not single requests.

How it uses Jev

It sends packed requests to Jev, asking a typed question per item (e.g. "Does this message need a human to act on it today?"). Jev returns calibrated probabilities; the library exposes them as answers with a .yes flag and .item, used to filter or route rows in code.

Primitives:choice

Technique worth stealing

Pack up to 32 items per request so one unit of a request-counted rate limit covers many judgements.

Try it

python demo.py (no key needed); pip install "jev-ultralightspeed[fast]"

View on GitHub

judged by Jevjev-1.13.0

Evidence

Each line is one question put to Jev about the README. ≥ 0.60 reads as yes, ≤ 0.40 as no; in between Jev is not making a call.

  • Jev-centricyes0.85
  • Shows a System One patternyes0.80
  • Handles uncertaintyno0.05
  • Measuredyes0.62
  • Runnableyes0.82
  • Worth recommendingno0.39
  • Model replicano0.09
  • Problem scopescore on a 0–2 scale1.98
  • About Jevyes0.96

Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 21 Sept 2026.