The public analytics agent benchmark launched Sept 1 2026, and it really resonated online. One of the most commonly asked questions was some version of “How are you running these?“. So I did a write-up on how we measure accuracy and how we keep moving pieces (data / semantics and LLM) as apples to apples as possible.
1 comment
[ 4.5 ms ] story [ 13.0 ms ] thread