Nearly nine years ago, three childhood friends began studying together at the Technical University of Munich. They parted ways after graduating: Rick Schneider went to Paris to scale up the GTM team at Celonis, Maxime Allard went to New York for grad school in Computer Science at Columbia followed by Imperial College where he earned his PhD, and Mathieu Klop continued in Munich earning his MD after which he became a practicing physician. In 2024 the team reunited, fueled by their shared past and set in motion by a common mission. The three were in agreement that accelerating drug discovery with AI was one of the most pressing challenges we face today and decided to found Helical to make their mark on this problem.

The premise of using AI to accelerate drug discovery holds immense potential. Today roughly 50 new drugs are approved by the FDA every year but there are thousands of known diseases that may be addressed. While each country has a different protocol to progress drug candidates for approval the insertion points for foundation models have become more clear. Helical decided to focus on two distinct activities: (1) early discovery — including target identification and biomarker discovery — and (2) in-silico experimentation at scale.

Foundation Models have demonstrated off-the-shelf value to digital tasks such as software engineering and customer service. The landscape of Bio Foundation Models and their constituent capabilities however presents a distinct problem — fragmentation. There exists a model maze that every organization must traverse to augment their capabilities. Helical’s platform is the unifying interface which introduces a research lab to the ideal model for their workflow and drug program. As the team likes to say — biology is too complex for a single model to rule them all.

Due to the diverse set of tasks and compatible models, life science researchers can end up with the task of managing a broad set of integration points, evaluation frameworks, incompatible embedding spaces, and infrastructure needs. Helical helps teams focus on pushing forward research instead by issuing (1) a unified interface for best-in-class open-source and proprietary DNA, RNA & single-cell Foundation Models, (2) a personalization platform for teams to align and post-train these models and (3) an experimentation platform for teams to inference models on their data across a set of common statistical tests.

Helical has quickly progressed to production deployments at several top-20 pharmaceutical companies with customers including Pfizer and Tanabe. The product has begun to reorient how some of the best research teams in the world approach critical tasks within early stage discovery such as target identification and biomarker discovery. In two short years hundreds of models have been fine-tuned with Helical and their platform has generated tens of billions of predictions at 1000x throughput year over year.

We’re excited to back Helical as part of their $10M seed round alongside redalpine, Frst, and existing investors including Aidan Gomez & Ivan Zhang (Cohere), Clément Delangue (Hugging Face), Mario Götze, and BoxGroup.