Unsupervised Biological AI
What if you could train a biological foundation model on... the entire internet? Exonic is pioneering a new generation of biological foundation models, focused on heterogeneous, unstructured, and noisy datasets. Our first application is the design of safer gene therapies. In 2025, we used AI to set a new state of the art in liver cancer targeted gene therapy, validated in vitro in our lab in San Francisco. So far in 2026, we have trained a new model with unprecedented zero-shot generalization on genomic regulatory expression in hold-out biosample datasets. More to be shared soon.
Founder
Co-founded Free2Grow, one of the largest AI voice companies in the home services vertical today. Co-led Turnpoint contract ($1.4B revenue aggregator) at Free2Grow. Columbia Computer Science and Mathematics, 2020. Published AI researcher (EMNLP, 2023). Led machine learning at Long Run Partners, multi-hundred million dollar NYC mortgage bank (employee #9). Worked on ML team at Numerai, the crowdsourced hedge fund ($500m+ AUM).
