Company: Bioptimus
CEO and Co-founder: Jean-Philippe Vert
Website: www.bioptimus.com
About Bioptimus
Founded in 2023 in Paris, Bioptimus is building a first of its kind AI foundation model platform for biology. The company was founded on the idea that biology cannot be fully understood through fragmented systems and isolated datasets. Biology operates as a connected system across genes, cells, tissues, and patients, yet most AI tools and research workflows still analyse these layers separately. As a result, drug development remains slow, expensive, and highly unpredictable.
Bioptimus develops AI foundation models that learn patterns in biology in the same way large language models learn from text. Rather than relying on hand engineered rules, the models use self supervised learning to discover biological relationships directly from data. The company was founded by researchers and AI leaders from Google DeepMind, Google Brain, Owkin, and leading academic institutions with expertise spanning machine learning and computational biology.
One of the biggest challenges Bioptimus faced early on was access to the scale and quality of biological data needed to train foundation models. To address this, the company built a large multimodal biology data infrastructure through partnerships, licensed datasets, and proprietary data generation initiatives.
Its first model, H-Optimus, has become a leading pathology foundation model with more than one million downloads worldwide and thousands of users across academia and industry. H-Optimus is used by leading pharmaceutical companies and research institutions, with collaborations including Proscia, AWS SageMaker, and MIT for clinical cancer research and prediction. The model has helped establish foundation models as a new paradigm for computational pathology and biomedical AI research.
Building on this momentum, Bioptimus introduced M-Optimus, a multimodal and multiscale World Model of Biology that integrates tissue, molecular, and clinical data within a unified AI system. To support the continued development of M-Optimus, Bioptimus launched STELA, a multinational initiative aiming to profile up to 100000 patient specimens across five continents and build one of the world’s largest clinically linked spatial biology resources.
By combining frontier AI research, large-scale proprietary data generation, and broad real world adoption, Bioptimus is helping establish a new foundation for biomedical AI and precision medicine.
Want to find out what TechRound can do for your business?
