Introducing Exa's Company Search Benchmarks

What this post added

Introduced a new benchmark for company search, including an ~800 query dataset and an open-sourced evaluation harness. This benchmark focuses on retrieval and RAG tracks, with specific handling for static vs. dynamic company attributes (founding year, location, industry, employee counts, funding history). The dataset was designed to avoid well-known companies to test retrieval capabilities rather than LLM memorization. The post details query types for attribute filtering, funding, composite, and semantic queries, as well as RAG queries for extracting static and dynamic facts.