AI research and products that learn from every interaction to power every application

We care about building a single core model for understanding human behavior.

Human behavior unfolds over time. Decisions, actions, and responses form sequences that reveal deeper patterns than any snapshot could capture. Recent advances in sequential modeling and self-supervised learning make it possible to learn from these patterns systematically, across domains.

At the heart of our approach is a simple insight: actions speak louder than words. An agent, whether a person, team, or system, is best understood by how it acts. An item, a product, document, or resource, is best understood by how others engage with it. These sequences of interaction become our foundation for learning.

We strongly believe that focus should lie in building a good model for the core understanding of agents and items, because it will be helpful in any application where either agents or items are involved; all such applications should be seen as downstream or fine tuning tasks, rather than problems that should be attacked in isolation.

We think this matters for a practical reason: any market defined by shared behaviors and meaningful connections should benefit from a model trained on all of its data together. A unified approach captures relationships that fragmented systems miss, and positions that model to serve any downstream need where understanding agents or items is valuable.