Rank #765 · on radar since 2026-07-03

LLM-Regression-Guard

End-to-end MLOps pipeline that catches LLM quality regressions before production. Every PR is scored against a versioned golden dataset using BERTScore + ROUGE-L + an LLM-as-Judge rubric, compared to the MLflow production baseline, and shadowed against 5% of live traffic. FastAPI + Celery + TimescaleDB + Streamlit + DVC + GitHub Actions.

Why it's ranked

Every score decomposes into published factors — the same math for every tool, paid or not. Read the methodology →