Owkin is a TechBio company that uses AI and federated learning to analyze decentralized patient data, aiming to improve drug discovery, clinical trials, and diagnostics. Pulse 2.0 interviewed Owkin co-founder and CEO Dr. Thomas Clozel to learn more.
Dr. Thomas Clozel’s Background
Could you tell me more about your background? Dr. Clozel said:
“I am a medical doctor by training, originally trained in Paris and New York, and I spent time as an assistant professor in clinical oncology.”
Formation Of The Company
How did the idea for the company come together? Dr. Clozel shared:
“When I was a doctor I saw how advanced our technology is getting – but how slowly it translates into benefit for patients. That’s how Owkin came together. It’s clear that the current pharma system for drug development is not working.”
“I think that’s because biology is too complex for the human mind alone. You need patient data, AI and patient validation to get there. That’s why I co-founded Owkin in 2016 with Gilles Wainrib, who is a machine learning genius, while I brought the clinical perspective and a focus on real patient impact.”
Favorite Memory
What has been your favorite memory working for the company so far? Dr. Clozel reflected:
“I don’t know, there’s so many to choose from! One was when I saw the first microanatomy analysis models working, back near the beginning. And then when our machine learning model mesoNET predicted the overall survival of malignant mesothelioma patients from digital pathology images better than existing subtype classification used by pathologists. I thought – this can really work. That was amazing. And there’s so many individual memories with amazingly smart people that I’ll never forget.”
Core Products
What are the company’s core products and features? Dr. Clozel explained:
“Owkin is building an agentic AI infrastructure to understand complex biology with a view towards helping pharma. Unlike other AI based drug discovery efforts that use lab and brokered data, this model is powered by multimodal patient data from our massive network of scientists and partners (including 800-plus hospitals), and from work in our lab. The infrastructure has yielded these products:
- K Pro – our Agentic AI scientist for smarter biopharma decision making. We license it to biopharma companies to analyse and understand their data and the literature and plan follow-on experiments and trials.
- Our unparalleled data network allows us to train the best AI tools and skills to power K Pro.
- Our sister company, recently spun out from Owkin Dx, called Waiv, produces diagnostics products, including AI diagnostics such as RlapsRisk BC and MSIntuit CRC, for breast cancer and colorectal cancer, respectively.”
Challenges Faced
Have you faced any challenges in your sector of work recently? Dr. Clozel acknowledged:
“One ongoing challenge in this sector is moving from ‘AI that predicts’ to ‘AI that explains and generalizes,’ especially in biology where context matters and datasets are heterogeneous. This AI reasoning is a real holy grail of the field. We’ve tackled that by grounding our models in real patient data, and turning our high-value proprietary datasets (like MOSAIC – the largest spatial dataset in oncology) into creative new Q&A datasets that allow us to help LLMs to understand how to reason better. But we’re still innovating here and there’s lots of new stuff in the pipeline – we believe that the team who can crack causality in biological systems first will win the race for what we call “biological artificial superintelligence,” or BASI.
Evolution Of The Company’s Technology
How has the company’s technology evolved since launching? Dr. Clozel noted:
“Enormously. We’ve gone from using representation learning, pulling information from microanatomy slide images, to incorporating multimodal data across different scales of biology – with the aim of pulling out deep underlying biology. Obviously AI has changed massively in the last 10 years and our stack has changed with it, incorporating foundation models, reinforcement learning and agentic infrastructure, for example. But something we’ve never changed is our focus on patient data as the bedrock of all of this.”
“That said, the market demand has changed over time – moving from Pharma companies that rely on AI services to those with their own AI team who want AI tools that give them the autonomy to work on their own data. So we’ve decided that the best play is to shift from a pure pharma services model to packaging all our technology into a single AI Scientist that can supercharge pharma productivity and pipelines.”
Significant Milestones
What have been some of the company’s most significant milestones? Dr. Clozel cited:
- Building a large-scale patient data network from tens of millions of patients and crucial relationships with 156 KOLs
- Founding MOSAIC, the world’s largest spatial multi-omics atlas (showing the integrated molecular layers) for cancer research across nine indications.
- Publishing 80 scientific papers in leading journals like Nature and Nature Comms.
- Launching validated diagnostic products (for example, RlapsRisk BC and MSIntuit® CRC) and spinning out our diagnostics division into Waiv.
- Incubating Bioptimus – a world-leading biology foundation model company.
- Launching our AI-enabled, AI-enabling, INVOKE phase 1a/1b clinical trial.
- Scaling the company with major strategic partnerships and financing, and becoming a unicorn.
Customer Success Stories
Can you share any specific customer success stories? Dr. Clozel highlighted:
“We have successful projects with major pharmaceutical companies, including Sanofi and BMS, and major tech companies Anthropic and NVIDIA – but I’ll refer you to the press releases here. We also partner and drive value for 104 leading academic institutions.”
Funding
Are you able to discuss funding? Dr. Clozel revealed:
“On previous funding, yes: Owkin has raised over $300M from leading biopharma companies and venture investors.”
Total Addressable Market
What total addressable market (TAM) size is the company pursuing? Dr. Clozel assessed:
“The company operates in a large and rapidly growing market at the intersection of pharma, biotech, technology, and healthcare providers. While we don’t disclose a specific TAM figure at this stage, the opportunity spans multiple sectors and geographies.”
Differentiation From The Competition
What differentiates the company from its competition? Dr. Clozel affirmed:
- Ground truth from real patient data at scale, built through a hospital partner network, not only curated lab datasets.
- A focus on causal and mechanistic understanding of biology.
- A full stack approach that connects AI, biology, and real-world clinical context, including diagnostics and wet-lab feedback loop, to validate our AI models in real patients.
- 10-year headstart compared to our competitors to win the BASI race – Pharma AI, multimodal data and established KOL relationships”
Future Company Goals
What are some of the company’s future goals? Dr. Clozel concluded:
“Biological Artificial Super Intelligence (BASI)! We are building towards AI that can reason across the full complexity of biological systems. This will be a game changer, something that can actually understand biology in a way that humanity never has before. This is part of our AI scientist project, which we want to use to supercharge biological research.”
