Run our beta semantic-search through 50 representative queries, score the results, and tell us what to fix before launch.
Run 50 representative queries through our beta semantic search, score relevance against our rubric, and document the top issues with reproduction steps.