Jev in production › SQL and data pipelines

jev-ultralightspeed

Python client that packs up to 32 items into each Jev request to answer one question per row across a dataset. Over 30,000 judgements against human labels it ran for 41% less money than one request per item, with no detectable accuracy difference (author).

Open on GitHub ↗

41% less moneymeasured against a baseline, as published by the source
Use
SQL and data pipelines
Industry
Data and analytics
Form
Open-source tool
Stage
In production
Listed
2026-09-21
Found via
discord
Repository
collapseindex/jev-ultralightspeed
Stars
12
Forks
0
Last push
2026-09-22
Language
Python
License
none stated

The README opens with

You have a pile of text and one question about each. Fifty thousand support tickets to triage. A quarter of reviews to sort by sentiment. A month of logs to flag. A column to backfill on a table you already have. This answers the question for all of them, with TypeSafe's Jev, in the time it takes to get a coffee.

32x the throughput of one request per item, for 41% less money, with no accuracy difference this benchmark can detect. The 32 is the pack depth, and under a ceiling counted in requests that is arithmetic rather than a measurement. The 41% and the accuracy are measured, over 30,000 judgements against human labels, with a script in this repository.

Badge

For the project's own README, linking back here:

Listed in Jev in production

Also used for sql and data pipelines