Systems Agents Discovery

Projects

A focused stack: cognition, agents, and the machinery to run discovery cycles.

H-OS

A cognitive operating system that structures reasoning into discovery cycles: hypothesis → simulation → evaluation → memory integration.

Status Experimental

Neo Agent

An autonomous agent with planning, episodic memory, tool use, and long-term learning loops. Designed for real-world experimentation and future embodiment.

Focus Long-horizon work

Discovery Engine

Generates hypotheses, runs simulations, evaluates outcomes, and updates knowledge graphs to accelerate research across domains.

Output Verifiable insights

Simulation Stack

A modular simulation harness for parameter sweeps, test-time compute experiments, and repeatable evaluation pipelines.

Goal Repeatability

Want to collaborate?

We work with partners who care about rigor, interpretability, and real experimental velocity.