Converge Bio Secures $25M Series A for AI Drug Discovery Evolution

A single iteration can now boost protein yield by more than four times. Converge Bio, a startup headquartered in Boston and Tel Aviv, has raised $25 million in an oversubscribed Series A round. Led by Bessemer Venture Partners, the funding also saw participation from high-profile executives at Meta, OpenAI, and Wiz, signaling intense interest in AI's potential to overhaul the pharmaceutical R&D timeline.

Why Converge Bio AI Drug Discovery Series A Matters

According to TechCrunch, the company's platform integrates generative AI trained on DNA, RNA, and protein sequences directly into existing research workflows. Since its inception just two years ago, the team has expanded from nine to 34 employees and signed 40 partnerships with global pharma and biotech firms. This rapid scaling follows a $5.5 million seed round in 2024, showcasing the industry's shift from skepticism to eager adoption.

The platform's strength lies in its specialized systems, such as antibody design and protein yield optimization. In one case study, a partner achieved a 4 to 4.5X increase in protein yield. CEO Dov Gertz noted that the industry is moving away from traditional trial-and-error toward more predictable, data-driven molecular design.

Sponsored

Advertise with Us

[email protected]

Sponsored

Combatting Hallucinations with Molecular Precision

While text-based LLMs often hallucinate, the stakes are far higher in biology where validating a molecule takes weeks. To mitigate this risk, Converge Bio pairs generative models with predictive filters and physics-based docking systems. This layered approach ensures that the antibodies or proteins generated are not just novel but functionally viable in three-dimensional space.

Addressing skepticism from AI leaders like Yann LeCun, Gertz clarified that they don't rely on text-based models for core science. Instead, they use specialized architectures like diffusion models and statistical methods tailored to biological data. The goal is to become the 'generative lab' for every life-science organization, complementing traditional wet labs with high-speed computational hypotheses.