Hook panel A/B tester

1,800-point thread debating whether typed decisions replace LLM calls for classification, routing, and scoring.

One direct Jev question per row against 12–14 Jev-scored dimensions with locally fitted weights on three classification tasks: 5,477 test rows, 25,174 Jev calls, $1.43. Decomposition wins on Japanese NLI (0.9076 vs 0.8373) but flags about 25× more hard benign rows as attacks (37.2% vs 1.5%).

DuckDB extension that classifies rows in CSV, Parquet, or DuckDB tables with Jev, about 10 seconds per 1,000 rows.

Speculative fan-out, confidence-gated routing, composite scoring, and intent routing.

Evidence-backed index of real-world Jev (TypeSafe AI System One) use cases: repos, patterns, benchmarks, and measured results

A curated list of public projects, integrations, and discussions built on Jev — TypeSafe AI's System One model for typed decisions.

A verified, community-maintained catalog of 485 open-source projects built with Jev.

Company homepage, waitlist, and product overview.

Introduction, primitives, patterns, cookbooks, HTTP API, and SDK references.