Olam LabsWork with us
We’re deeply familiar with the formats, standards, and types of tasks that teams at the frontier of training and research are looking for.
Pre-deployment evaluations
Pre-deployment evaluations for safety and capability through anonymized arena models and internal evals.
Datasets
Tasks and datasets to improve models in Multi-Agent Arena in both capability and safety.
Real-world environments
Real-world enterprise tasks for code, computer use, game dev, etc. to improve model capability and safety through multi-agent environments.
Custom environments
We’re specialists and at the frontier of both RL and evaluations for multi-agent environments like swarms or social behavior.
Get in touch
Talk to our team about how we can help you in your RL, research, and evaluations.