Proxima $80M funding fuels AI drug design for proximity therapeutics

Proxima, the AI-native biotech formerly known as VantAI, has secured an oversubscribed $80 million seed financing round, led by DCVC and featuring NVentures, NVIDIA’s venture capital arm. The Proxima $80M funding is a significant bet on the next frontier of drug discovery: proximity-based medicines.

These next-generation therapies, which include modalities like molecular glues and PROTACs, operate by controlling how proteins interact with one another, rather than simply inhibiting a single target. This approach is crucial for addressing the vast majority of disease targets—estimated at over 95% of the human proteome—that are currently considered "undruggable."

The company rebranded to Proxima to reflect its singular focus on building the technological and data infrastructure necessary to unlock this field. CEO Zachary Carpenter noted that progress in proximity therapeutics has been constrained by a severe lack of structural data and accurate design tools.

Mapping the Undruggable Proteome

Proxima’s core platform combines its NeoLink data-generation technology, which produces structural information on protein complexes at a proteome-wide scale, with its Neo AI model series. This pairing creates an end-to-end pipeline designed to rationally design proximity-modulating small molecules, aiming to shorten development timelines and improve safety.

The investment will accelerate Proxima’s internal pipeline and its existing multibillion-dollar collaborations with major biopharmaceutical companies, including Johnson & Johnson and Bristol Myers Squibb. With programs already advancing toward the clinic, the first co-developed therapies are expected to enter clinical trials by 2026.

DCVC General Partner Jason Pontin, who joins the board, highlighted that Proxima’s combination of proteome-scale structural data and state-of-the-art generative AI foundation models positions the company uniquely to discover entirely new classes of medicines. The participation of NVentures underscores the growing recognition that frontier AI models are now essential infrastructure for solving complex biological challenges.