JevMade hello@JevMade.com
← Back to experiments

Benchmarks & research

padflow-jev-evals

PadFlow contributes anonymized land-development decisions for testing confidence-aware models such as Jev.

Source screenshot of padflow-jev-evals
SOURCE SCREENSHOT · source ↗ · captured 2026-09-21Full screenshot ↗

What it does

The repository includes task schemas, labeled rows, a runner, and calibration checks for document routing and transaction coding.

Maker-reported (not independently measured by JevMade): model that is 90% accurate and always says 0.95 · n = 5 / 5 / 7 rows per decision (route_document / code_transaction /

Primitives
score
Platform
Python
Added
Project created
GitHub stars
1 · snapshot 2026-09-18T22:58:45Z

Source checked 2026-09-19 — opened the GitHub repository directly.