Fulcrum Research

Much of the work of the future will be executed by AI agents. They will write software, perform research, and carry out knowledge tasks we can't yet imagine.

Our current stack for understanding is not ready for a world that humans didn't build. We want to scale human oversight with model intelligence.

Currently, we are building tooling that tells you what your ML evals and environments are truly testing for. In the long run, we want to build the inference time infrastructure to safely deploy agents in the world.

If you are building RL environments, evals, or deploying agents, contact us to try our tooling.

You can reach us here: team@fulcrumresearch.ai.