“Bioptimus Secures $35M in Seed Funding to Propel AI-Driven Biological Model Development”

A new generative AI startup has emerged in Paris, France – Bioptimus. But what sets them apart is their exclusive focus on utilizing the collective knowledge we have garnered in AI over the years, specifically in the field of biology.

The decision to solely concentrate on biology was a strategic one for Bioptimus, as the acquisition of training data in this area is not as straightforward. While OpenAI is moving away from web crawling and shifting towards licensing agreements with content publishers, Bioptimus faces different challenges with sensitive clinical data that is not publicly accessible.

Similar to other AI startups, Bioptimus will require significant capital investments to train their models on expensive GPUs and hire top researchers. To support their endeavors, the startup has raised a seed round of $35 million, led by Sofinnova Partners. Additional investors such as Bpifrance’s Large Venture fund, Frst, Cathay Innovation, Headline, Hummingbird, NJF Capital, Owkin, Top Harvest Capital, and Xavier Niel have also contributed to this funding round.

Bioptimus may be a new player, but it is not coming out of nowhere. Led by co-founder and executive chairman Jean-Philippe Vert, the company’s day-to-day operations will be managed by someone else. Jean-Philippe Vert is also the Chief R&D Officer at Owkin, a French biotech unicorn that leverages AI to identify new drugs and enhance diagnostics.

Rodolphe Jenatton, the CTO of Bioptimus, brings additional AI experience to the team, having previously worked as a senior research scientist at Google. Several other co-founders are also former Google DeepMind researchers.

Bioptimus will benefit from Owkin’s collaborations with top biopharmaceutical companies, gathering multimodal patient data from leading academic hospitals worldwide. This valuable dataset will serve as the backbone for training their foundational model.

A Massive Undertaking by Owkin

Bioptimus could potentially be seen as a spin-off company or a moonshot project from Owkin. Why didn’t Owkin pursue this foundational model in-house instead? The creation of new AI models is such a monumental task that a separate entity seemed like a more logical choice.

“Developing foundational models in biology is not on Owkin’s immediate roadmap. However, Owkin is supportive and willing to collaborate with a company like Bioptimus. The training of large-scale foundational models requires significant resources in terms of data volume, computing power, and diverse data modalities. These resources are more readily accessible to a specialized entity, like Bioptimus,” stated Jean-Philippe Vert to TechCrunch. “As a ‘pure player’ in foundational models, Bioptimus is better suited for this endeavor.”

The startup has also established a partnership with Amazon Web Services, indicating that their model will be trained in Amazon’s data centers. With ample funding, it’s now time for Bioptimus to focus on developing their AI model and see what breakthroughs can be achieved in the biotech research community.

“Our goal is to ultimately utilize our AI to advance disease diagnosis, precision medicine, and aid in the development of new biomolecules for medical and environmental purposes,” expressed Vert.