Bespoke Labs
About
Bespoke Labs builds production-like RL environments and infrastructure for training, evaluating, and optimizing AI agents. Its work spans long-horizon simulations, agent benchmarks, open data, and optimization systems such as GEPA.
What They Offer
Commercial or operational services this company provides.
Agent evaluations
Rubrics / verifiers
Open-source tooling
Custom RL environments
Training datasets
Managed infrastructure
Products & Public Artifacts
OpenThoughts
Dataset · Open source
Open reasoning dataset distributed on Hugging Face.
Visit artifact →Terminal-Bench
Benchmark · Public
Benchmark for evaluating AI agents on terminal-based tasks in realistic execution environments.
Visit artifact →GEPA
Framework · Public
Optimization approach for improving AI systems through reflective prompt and system evolution.
Visit artifact →Technical Capabilities
Long-Horizon TasksCode / Terminal ExecutionProgrammatic VerifiersReal-App / Workplace SimulationSandboxed Execution
Only publicly documented capabilities are shown. Absence does not imply the capability is unavailable.
Focus Areas
CodingRL EnvironmentsTraining DataRLHF / Post-trainingEvaluations
Domains
Software EngineeringEnterprise
Leadership
Mahesh Sathiamoorthy
Co-founder
Alex Dimakis
Co-founder