Building multi-agent simulations for model evals and training.
Multi-agent environments let us evaluate and train models in complex, simulated worlds. We work with researchers on both evaluations and training for character and agentic performance typically difficult to assess for using current popular datasets. Play one of our first releases, Multi-Agent Arena (https://olamlabs.ai/arena), where humans come play social strategy games against multiple AI agents. It's currently top 50 on OpenRouter and has thousands of matches played each day. We use Multi-Agent Arena to build datasets on agentic performance and real-world socialization.
Founder/CEO
I am building the entire stack for multi-agent environments so we can do better AI evaluations and training. I'm a tech enthusiast and optimize for following my curiosity. In the past I've done some fun things before Olam Labs ranging from esports, to consumer, to engineering, to research. I was one of the best Valorant players in the world, a Staff Engineer at AI Dungeon, done math research, and other stuff. You can read more here: https://www.sensho.xyz/
Founder/CTO
currently building mulit-agent envs to create better evals @ olam labs, previously: was doing a cs/math at waterloo and business double degree at laurier (dropped out to start olam labs ) creating reinforcement learning systems and doing data science @ RBC did algorithmic arbitrage on prediction markets briefly top 50 on fifa mobile
