Converge Bio raises $25M, backed by Bessemer and execs from Meta, OpenAI, Wiz

The pharmaceutical and biotech sectors are increasingly embracing artificial intelligence to streamline drug discovery processes, aiming to shorten research and development timelines and improve success rates amid escalating costs. In this rapidly evolving landscape, over 200 startups are vying to integrate AI into research workflows, attracting significant investor interest. Converge Bio, a startup based in Boston and Tel Aviv, has emerged as a key player by securing $25 million in an oversubscribed Series A funding round, led by Bessemer Venture Partners, with participation from TLV Partners, Vintage Investment Partners, and additional backing from unnamed executives at Meta, OpenAI, and Wiz. This funding comes at a time when competition in AI-driven drug discovery is intensifying. The startup leverages generative AI trained on molecular data to assist pharmaceutical and biotech companies in accelerating drug development. Converge's innovative approach involves training generative models on sequences of DNA, RNA, and proteins, integrating these models into the established workflows of its clients to expedite the drug development lifecycle. CEO and co-founder Dov Gertz emphasized the importance of their platform, stating, “Our system supports various stages of drug development, from target identification to clinical trials, helping bring new therapies to market more rapidly.” Converge Bio has already introduced three specialized AI systems: one for antibody design, another for optimizing protein yields, and a third for discovering biomarkers and targets. Gertz elaborated on the antibody design system, explaining, “It consists of three interconnected components: a generative model that creates novel antibodies, predictive models that evaluate their molecular properties, and a docking system that simulates the interactions between the antibody and its target.” He noted that this comprehensive system provides clients with ready-to-use tools that seamlessly integrate into their existing workflows. This latest funding round follows a $5.5 million seed round raised in early 2024. In just two years, Converge has rapidly scaled its operations, forming 40 partnerships across the U.S., Canada, Europe, and Israel, and is now expanding into Asian markets. The team has grown from nine to 34 employees within a year, reflecting the startup's accelerating momentum. Converge Bio has started to publish case studies demonstrating its impact; one highlighted a partner's ability to increase protein yield by 4 to 4.5 times in a single computational iteration, while another showcased the platform's capability to generate antibodies with exceptionally high binding affinities in the single-nanomolar range. Interest in AI-driven drug discovery is surging, with notable collaborations like Eli Lilly's partnership with Nvidia to create a supercomputer for drug research and the recent Nobel Prize awarded to the creators of AlphaFold for predicting protein structures. Gertz remarks that the current environment presents unprecedented financial opportunities in life sciences, as the industry transitions from traditional trial-and-error methods to data-driven molecular design. Despite initial skepticism when the company was founded, Gertz notes that this has shifted dramatically, driven by successful outcomes from Converge and academic research. However, challenges persist, particularly regarding the use of large language models (LLMs) in drug discovery. Gertz points out that while LLMs can analyze biological sequences, their reliability can be problematic due to issues like hallucinations, which are harder to identify in molecular contexts than in text. To mitigate these risks, Converge employs a combination of generative and predictive models to filter new compounds, aiming to enhance success rates for its partners. Gertz acknowledges the critiques from experts like Yann LeCun, agreeing that while LLMs are not central to their scientific framework, they can serve as supportive tools for navigating literature on generated molecules. Converge Bio envisions a future where every life-science organization utilizes its platform as a generative AI lab, complementing traditional wet labs with computational capabilities to create hypotheses and molecules. “We aspire to be the generative lab for the entire industry,” Gertz concluded.