Building intelligence from nature’s first principles
We’re a biocomputing company deriving biological AI architectures from living nervous systems and developing brains.
A control architecture derived from mapped biological structure
We’re reconstructing how biology builds intelligent systems.
Brains do not learn from a blank slate. Genome, development and neural activity build structured circuits before experience begins, giving biological systems capabilities that today’s AI must largely learn from data.
We combine connectomes, gene expression, neural development and living experiments to recover the rules that generate those architectures, then rebuild them as computational models. As we move across increasingly complex nervous systems, the same loop produces more capable models while expanding the biological systems in which those architectures can be tested and instantiated. Over time, this becomes a new way to build intelligence across software and biology.
Our first commercial model launches in December, designed for generalization, sample efficiency and continuous adaptation in changing environments.