Olam Labs
Building multi-agent simulations for model evals and training.
About Olam Labs
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.
Founders
Om Buddhdev
Co-Founder
I want to build rly cool simulations At 11, I was posting YouTube videos about building PCs and hardware projects At 12, I was one of the wealthiest players on Minecraft's largest economy sim server At 13 & 14 I was making Roblox games played millions of times At 15 & 16 I was competing in Fortnite, peaking at top 100 NA. At 17 & 18, I was viral on TikTok with 195k followers. At 19 & 20, I was 8x radiant and peaked top 25 in Valorant At 21, I was a Staff Engineer at AI Dungeon / Latitude
shreshth sharma
Founder
currently building simulated envs to create better evals @ olam labs, previously: was doing a cs and business double degree at uwaterloo (dropped out for yc) building reinforcement learning systems and doing data science @ RBC i was briefly top 50 on fifa mobile