Open-source benchmark for evaluating LLMs on 220 real professional tasks across 9 sectors and 44 occupations. Reproducible experiments, artifact validation, grading, and a live evidence dashboard.
-
Updated
Jul 29, 2026 - Python