JevMade Sign in
← Back to experiments

Apps & data pipelines

jev-table

Bookmark: jev-table Keep this in your collection.
Leave a noteWhat would you try with this? : jev-table

Only you can see your notes.

Source screenshot of jev-table
SOURCE SCREENSHOTFull screenshot ↗

What it does

Label rows in a CSV or JSONL file with Jev, and set uncertain answers aside for a person to review.

How you can use it

Start by writing down the questions you need answered for each row of your spreadsheet. For example, you might ask whether incoming text messages are spam or how quickly someone must reply. A developer can then set up the tool using an access key from TypeSafe or your own office server.

The program marks doubtful answers in a separate review column so you can check and fix them by hand. It can estimate costs before handling your data, but running files with more than ten thousand rows requires an extra safety confirmation.

Maker-reported (not independently measured by JevMade): Live smoke on 200 UCI SMS rows with jev-1.13.0: 97.4% accuracy on 194 comparable rows, all 93 auto-accepted rows correct, 46.5% automated, $0.0040, 8.9 s (author-reported) · About 477 input tokens per row; per-call latency p50 4.7 s and p95 8.4 s; dry run 91,871 tokens against 95,313 actual (author-reported)

Primitives
choice, score, noul
Platform
Python
Added
Project created
GitHub stars
0 (snapshot, not live)

Keep this for later

Sign in to bookmark experiments, guides and videos, and keep notes only you can see.

Continue to sign in

We’ll bring you back to this listing.