Help us build the next generation of AI
New architectures are necessary for the next generation of AI.
Our founding team met as PhDs at Stanford AI Lab, where we invented State Space Models (SSMs) — a fundamental new primitive for training large-scale foundation models that are higher quality and more efficient.
Over the past four years, we've built the theory behind SSMs and scaled it up to achieve state-of-the-art results in modalities as diverse as text, audio, video, images, and time-series data.
Join our team
Our team is building at the edge of AI, pioneering new architectures for the next generation of intelligent systems.
Our values
Get the fundamentals right
The biggest wins come from foundational advancements and long-term thinking. We solve problems at the root, articulate the tradeoffs from shortcuts, and invest in infrastructure and process. Progress in AI models is fundamental to progress in AI.
Build to solve problems
We're accountable not just for how well we do our work, but for whether we're working on the right problems in the first place. We actively seek out the biggest bottlenecks and solve them. We build with our users in mind, and we work to understand our users and the reality they inhabit.
Aim beyond what's possible
We find it energizing to push beyond what the world thinks is possible. We aim to be the best in the world in everything we do. We continually iterate because we can always be better.
Be honest, with ourselves and others
We admit what we don't know, and we actively question our assumptions to try to do things better. We're always curious about how we can improve, and we're curious rather than defensive when we disagree. We're honest with others and say what we mean plainly and directly.
Have fun with it
We take the work seriously, but not ourselves.
Backed by world-class investors
We partner with leading investors in our field. Our backers include top venture capital firms and innovative entrepreneurs with strong product expertise.
