Applied Geometry is an AI research lab building biological superintelligence: AI that can make discoveries in biology and health beyond the reach of the best human research teams. Our goal is to put more intelligence behind a single problem than any institution can assemble today, and to compress years of discovery into days.
We are training frontier models to do science across biomedicine, building on methods that have driven extraordinary advances in coding and mathematics. Biology and human health present a harder frontier, where outcomes are less readily verifiable, tasks are longer-horizon, suitable training data is scarcer and living systems introduce complexity that resists simple rules or definitive answers. Progress requires models that can reason from incomplete evidence, choose experiments that distinguish between competing hypotheses and revise their conclusions when results contradict them. We give agents the tools and data to undertake this work, then use the outcomes to improve the models themselves, with the ambition to make understanding and curing disease one of the defining achievements of artificial intelligence.
The capabilities we are developing span biology, drug discovery and clinical medicine, from uncovering disease mechanisms and designing experiments to developing new therapies, improving diagnosis and guiding treatment decisions for individual patients. As agents master harder problems, we expand the scope of their work and train them to sustain coherent research over longer periods. We are building towards systems that can originate and pursue research programmes whose discoveries exceed what their creators could have anticipated.
Swarms of superhuman AI researchers will navigate biological search spaces far beyond the reach of any human institution, exploring vast numbers of hypotheses in parallel to uncover disease mechanisms, identify new drug targets and design treatments. This will accelerate discovery across a breadth of diseases no human research effort could pursue at once, opening paths to cures that have eluded generations of scientists.
We are assembling a world-class team to build one of the most consequential companies in the world, with the independence to pursue original ideas and the intensity to turn them into working systems.
Superintelligence must be measured by its consequences for human life, and we will not have achieved it until it transforms our ability to understand and cure disease. We are training frontier models to do the work that makes that future possible.
Build with us