LLM Agents
LLM Agents
LLM Agents
Building AI that learns like humans
The human brain is a sponge. Today’s AI brains are brittle and rigid. At Letta, we’re building self-improving artificial intelligence: creating agents that continually learn from experience and adapt over time.
Founded by the creators of MemGPT from UC Berkeley’s Sky Computing Lab (the birthplace of Spark and Ray). Backed by Jeff Dean, Clem Delangue, and pioneers across AI infrastructure. Our agents already power production systems, learning and improving every day.
research vision
Architecting the LLM OS
The large language model is the new CPU - a finite context compute unit, tokens in, tokens out. To build agents that can act in the world and learn from their experience, we need to build a new operating system that manages the context of the LLM, orchestrating a multi-tiered context hierarchy turning tokens into living memories.
Memory and Learning
Continual learning not only in weights, but also in token space
Continual learning not only in weights, but also in token space
Agentic Context Engineering
Enabling models to self-improve by editing their own context window
Enabling models to self-improve by editing their own context window
Sleep-time Compute
Scaling compute for offline learning so agents can think while they sleep
Scaling compute for offline learning so agents can think while they sleep
Featured papers
Join our team
We’re assembling a world-class team of researchers and engineers to solve AI’s hardest problem: making machines that can reason, remember, and learn the way humans do.
