
Drug discovery, the process of finding new molecules that can be used in drug development, is a notoriously time-consuming and expensive process. Traditional screening techniques, such as high-throughput screening, are not only costly but also have limited effectiveness, failing to meet the demands of drug development. However, a number of emerging biotechnology companies are attempting to accelerate and simplify this process using artificial intelligence and advanced data technologies.
Chai Discovery, an AI startup founded in 2024, is a prime example of such innovative companies. In just over a year since its founding, its young co-founders have successfully raised hundreds of millions of dollars, attracting the attention of several influential Silicon Valley investors and becoming one of the most closely watched companies in the field. Last December, the company successfully completed its Series B funding round, raising $130 million and reaching a valuation of $1.3 billion.
On January 9th, Chai further announced a collaboration with Eli Lilly, a pharmaceutical giant that will utilize Chai's software to aid in new drug development. Chai's Chai-2 algorithm focuses on developing antibodies—proteins essential for the body's defense against disease. The startup states that it is committed to creating a "computer-aided design suite" for molecular design.
The AI drug discovery field, in which Chai operates, is currently at a critical stage of development. Shortly after Chai announced its collaboration with Eli Lilly, Eli Lilly simultaneously announced a partnership with Nvidia, investing $1 billion to build an artificial intelligence drug discovery laboratory in San Francisco. This institution, called the "Joint Innovation Lab," will integrate big data, computing resources, and scientific expertise, with the core mission of accelerating the drug discovery process.
The AI drug discovery industry is not without its challenges, and there are still many dissenting voices. Some industry veterans believe that, given the immense difficulty of traditional drug development, these new technologies are unlikely to have a substantial impact. However, there are also many supporters in the industry, creating a balance of opposing viewpoints.
Elena Viboch, Managing Director at General Catalyst, told the media that the company firmly believes that companies using Chai's services will see tangible results. General Catalyst is one of Chai's main investors, and Vibhoh noted, "We believe that biopharmaceutical companies that partner with companies like Chai the fastest will be the first to advance drug molecules into clinical trials and develop truly important medicines. In fact, this means that by starting a collaboration in 2026, we could see the first novel drugs entering clinical trials by the end of 2027."
Aliza Apple, head of the TuneLab project at Eli Lilly, also expressed confidence in Chai's products. The TuneLab project focuses on leveraging artificial intelligence and machine learning to drive drug discovery. Apple stated, "By combining Chai's generative design models with Eli Lilly's deep expertise in biologics and proprietary data, we aim to expand the ability of AI to design better molecules from the outset, ultimately helping to accelerate the development of innovative medicines to benefit patients."
Although Chai has been in existence for less than two years, its origins can be traced back to about six years ago—when the company's co-founders had a crucial conversation with OpenAI CEO Sam Altman. One of the founders, Josh Meier, joined the OpenAI research team in 2018. After he left, Altman contacted Meier's college friend, Jack Dent, to explore potential business collaboration opportunities. Meier and Dent initially met in a computer science class at Harvard University, where Dent was working as an engineer at Stripe (Stripe was also one of Altman's early investments). Altman asked Dent if Meier would be willing to collaborate on starting a proteomics startup, a company focused on protein research.
Dent recalled that Altman "messaged me saying that everyone at OpenAI thought very highly of him, and asked if he would be willing to work with them to start a proteomics spin-off company." Dent immediately replied to Altman, "Of course," but there was a problem: Meier believed the technology wasn't mature enough at the time. The AI technology that such companies rely on—systems built on powerful algorithms—was still in its nascent stages and a long way from being stable and mature.
At the time, Meier was eager to join the Facebook research team, and he eventually did. During his time at Facebook, he participated in the development of ESM1, the world's first Transformer protein language model, and a key technological precursor to Chai's current research. After leaving Facebook, Meier worked for three years at Absci, another AI biotechnology company that also focuses on drug discovery. It wasn't until 2024 that Meier and Dent felt the time was right to launch the proteomics company they had initially discussed with Altman. "Josh and I reconnected with Sam and told him we should pick up where we left off – that we were going to start Chai together," Dent said.
OpenAI eventually became one of Chai's first seed round investors. In fact, Meier, Dent, and their co-founders Matthew McPartlon and Jacques Boitreaud officially founded Chai in the artificial intelligence giant's offices in San Francisco's Mission District. "They were incredibly generous in providing us with office space," Dent revealed.
More than a year after its founding, Chai is basking in the positive news of a new collaboration with Eli Lilly. Dent said the key to the company's rapid growth lies in building a high-quality team. "We've really put our heads down and pushed the limits of these models," Dent emphasized. "Every line of code in our codebase is proprietary. We're not just taking off-the-shelf large language models (LLMs) from the open-source ecosystem and fine-tuning them. These are highly customized architectures."
Viboch of General Catalyst told the media that she believes Chai is well-positioned for rapid deployment. "There are no fundamental obstacles to deploying these models in drug discovery," she said. "Companies will still need to test and clinically trial drug candidates, but we believe that companies that adopt these technologies will gain a significant advantage – not only in shortening development cycles but also in developing drugs that were previously difficult to develop."