From OpenAI to Eli Lilly: The Rise of Chai Discovery in AI Drug Development

The journey to create new medicines is undergoing a profound transformation, driven by artificial intelligence. Traditional drug discovery methods are notoriously slow and expensive, often relying on a trial-and-error approach that yields more failures than successes. A wave of innovative biotech firms is now harnessing advanced data technologies to fundamentally reshape this process, aiming to bring vital treatments to patients faster and more efficiently.

Among these pioneers is Chai Discovery, a startup that has captured significant attention since its founding in 2024. In just over a year, its founders have secured hundreds of millions in funding and attracted backing from some of Silicon Valley’s most prominent investors, catapulting the company to a $1.3 billion valuation after a recent Series B round. The firm’s momentum reached a new peak with a major partnership announcement involving pharmaceutical leader Eli Lilly. This collaboration will see Lilly utilize Chai’s proprietary software platform to aid in developing new therapeutic molecules.

Chai’s core technology, an algorithm named Chai-2, is engineered to design antibodies, the crucial proteins that combat disease. The company describes its platform as a sophisticated “computer-aided design suite” for molecular creation. This deal arrives at a pivotal time for AI in biotech, coinciding with Eli Lilly’s separate announcement of a substantial collaboration with Nvidia to establish a dedicated AI drug discovery laboratory. These parallel moves signal a strong industry belief in the potential of data-driven approaches to accelerate medical innovation.

Naturally, this rapid shift has its skeptics. Some seasoned professionals in the pharmaceutical field question whether new technologies can truly overcome the deep-rooted complexities of traditional drug development. For every doubter, however, there appears to be a committed believer with a compelling vision.

Elena Viboch, a managing director at General Catalyst, a key investor in Chai, expressed strong confidence in the startup’s trajectory. She believes that biopharma companies which quickly adopt such technologies will gain a decisive edge. “We believe the biopharma companies that move the most quickly to partner with companies like Chai will be the first to get molecules into the clinic, and will make medicines that matter,” Viboch stated. She projected that early partnerships could see novel medicines entering clinical trials by the end of 2027.

This sentiment is echoed within the industry. Aliza Apple, who leads Eli Lilly’s TuneLab AI program, emphasized the synergistic potential of the partnership. “By combining Chai’s generative design models with Lilly’s deep biologics expertise and proprietary data, we intend to push the frontier of how AI can design better molecules from the outset,” she explained, highlighting the shared goal of accelerating the development of innovative patient treatments.

The story of Chai’s origins is deeply intertwined with the broader AI ecosystem. While officially launched less than two years ago, its conceptual seeds were planted around six years ago in discussions between its co-founders and OpenAI CEO Sam Altman. Co-founder Josh Meier had previously worked on OpenAI’s research team. After his departure, Altman reached out to Meier’s former college friend, Jack Dent, then an engineer at Stripe, to explore a potential venture in proteomics, the study of proteins.

Altman inquired if Meier might be interested in collaborating on such a startup. Dent agreed it was a promising idea, but there was a significant hurdle: Meier felt the underlying AI technology wasn’t yet mature enough for their ambitions. He instead joined Facebook’s research team, where he contributed to developing ESM1, a pioneering transformer model for protein language that served as a crucial precursor to Chai’s current work. Following his tenure at Facebook and a subsequent role at another AI biotech firm, Meier felt the technological landscape had finally evolved sufficiently.

By 2024, Meier and Dent reconnected with Altman, ready to revive their original concept. They informed him they were launching Chai together. OpenAI subsequently became one of the startup’s earliest seed investors, even providing initial office space in San Francisco for Meier, Dent, and their co-founders to begin their work.

Reflecting on the company’s rapid ascent, Dent attributes their progress to assembling an exceptional team and a commitment to foundational innovation. “We really just put our heads down and pushed the frontier of what these models are capable of,” he said. He emphasized that their technology is built from the ground up, not simply fine-tuned from existing open-source models. “Every line of code in our codebase is homegrown. These are highly custom architectures.”

Viboch from General Catalyst sees this technical rigor as a key differentiator, positioning Chai for immediate impact. “There are no fundamental barriers to deployment of these models in drug discovery,” she asserted. While new drug candidates will still need to undergo rigorous testing and clinical trials, she believes early adopters will secure significant advantages. These benefits extend beyond simply shortening discovery timelines to potentially unlocking new classes of medicines that have historically been too challenging to develop.

(Source: TechCrunch)