We're building the infrastructure for a world where
every company owns its intelligence.
Our platform, AC2, lets teams train open models on their own tasks. They can then serve those models in production and feed production traffic back in as training signal. Our researchers work directly with customers to turn their data and expertise into models they couldn't get anywhere else. Teams like NVIDIA, Microsoft, Cognition, Harvey, DoorDash, and Mercor use AC2.
We're a team of about 70 people in San Francisco, and 85% of us are researchers and engineers. Many of us are former founders. Before Applied Compute we built RL infrastructure at OpenAI, data foundations at Scale AI, and systems at Together, Two Sigma, and Watershed. We're backed by Kleiner Perkins, Benchmark, Sequoia, Lux, Greenoaks, Elad Gil, NVIDIA, Blackstone, and others.
We're product-led, research-enabled, and forward-deployed. The people pushing on post-training research also sit with customers while their models go into production, and what we learn there goes straight back into the platform.
Like our models, we keep learning. People who do well here like hard technical problems that show up in the real world. They enjoy sitting with customers to understand how work actually gets done, and they've shown they can own something end to end, whether that's a company, a research project, or a long list of side projects.
Our benefits
We want you focused on the hardest problems in AI infrastructure, so we take care of everything else.
Don't see your role? We'd still love to hear from you.
Say hi