Research

Notes, results, and findings from our work on autonomous agent optimization.

  • Coming soon
    Beating τ-bench Outperforming the best public models on τ-bench retail — and what it taught us about agent optimization.