Causal Dynamics Lab is an AI research lab. We build products that give machines the understanding to act on complex systems.
Mission
Our purpose
Current AI systems optimize for sounding correct. They hallucinate confidently, forget what they learned yesterday, and burn enormous resources processing context they don't need.
Our mission is to build AI that reasons through cause and effect, learns from experience without forgetting, and grounding ourselves within the world model.
Our research spans causal inference, graph World Model, and reinforcement learning. We turn that research into products - starting with production software, the domain with the richest causal signal and the fastest feedback loops.
Roadmap
The ambition that sets us apart
We are pursuing three research frontiers simultaneously
01
01
AI that shows reasoning
Not post-hoc explanations bolted onto a black box. Causal structure built into how the system thinks. Humans could follow the chain, check the logic, and either trust or reject the conclusion.
02
02
AI that learns without forgetting
Current systems reset. Ours accumulates knowledge. An AI agent should get better over time, not start over every session.
03
03
AI that reasons on 20 watts, not 20 gigawatts
The human brain doesn’t process everything all at once. Reasoning starts with small, relevant subsets, and the rest is ignored. We're closing that gap with the infrastructure we have today.
Backed by world-leading experts and investors
Join us
We hire selectively across research, engineering, and forward-deployment roles. If you're drawn to work where the science and the product are the same thing, we'd like to hear from you.