Skip to main content

Proxima

Category: AI in Healthcare

An AI-native drug discovery platform specializing in induced proximity and geometric deep learning to design therapeutics for previously 'undruggable' protein targets. Proxima was founded in 2019. The company is led by Zachary Carpenter. Based in New York, USA. Team size: 11-50. Total funding raised: $80M. Latest round: Series B ($80M, Jan 2026 at $1.3B+ valuation) - AI sales intelligence; NOT Seed. Key investors include DCVC, NVentures (Nvidia), Braidwell, Roivant Sciences, AIX Ventures, Yosemite, Magnetic Ventures, Alexandria Venture Investments, Modi Ventures.

Founded
2019
Headquarters
New York, USA
Team size
11-50
Total funding
$80M

Value proposition

Accelerates the discovery of molecular glues and protein degraders by using 3D geometric AI to model complex protein-protein interactions, unlocking therapeutic pathways that traditional small molecules cannot reach.

Products and solutions

Proxima Design Platform (Foundation models for biomolecular interactions), Neo-1 Foundation Model (AI model for rewiring molecular interactions, launched Mar 2025), NeoLink Structural Proteomics Platform, Molecular Glue Discovery Engine, Targeted Protein Degradation (TPD) Suite, Geometric Deep Learning (GDL) Structural Models

Unique value

Unlike traditional AI drug discovery that focuses on single-protein binding, Proxima focuses on 'induced proximity'—designing molecules that act as bridges to bring two proteins together to trigger a biological response.

Target customer

Global pharmaceutical companies (Biopharma), biotechnology firms, and internal therapeutic pipelines.

Industries served

Biotechnology, Pharmaceuticals, Healthcare AI, Life Sciences

Technology advantage

Leverages Geometric Deep Learning (GDL) to treat proteins as 3D graphs/shapes rather than linear sequences, allowing for high-fidelity prediction of protein-protein interfaces and the design of highly specific 'molecular glues.'

How they differentiate

Proxima utilizes Geometric Deep Learning (GDL) to model proteins as 3D shapes rather than linear sequences, allowing for the de novo design of 'molecular glues' and induced proximity therapeutics for targets previously considered undruggable.

Main competitors

Monte Rosa Therapeutics, C4 Therapeutics, Neomorph, Kymera Therapeutics

Key partnerships

Nvidia (NVentures) - Strategic investor and compute partner, Bristol Myers Squibb (BMS) - Multi-year drug discovery collaboration, Janssen (Johnson & Johnson) - Research partnership in protein degradation, Boehringer Ingelheim - Strategic AI-driven discovery alliance, Roivant Sciences - Ecosystem partner and incubator, Halda Therapeutics - $1B+ strategic research collaboration for RIPTAC medicines (Aug 2025), Blueprint Medicines - Expanded collaboration for induced proximity drug discovery (May 2025)

Notable customers

Bristol Myers Squibb, Janssen (Johnson & Johnson), Boehringer Ingelheim, Blueprint Medicines (acquired by Sanofi), Halda Therapeutics

Major milestones

Founded in 2019 as VantAI within the Roivant Sciences ecosystem, Established strategic AI-driven drug discovery partnership with Janssen in 2022, Expanded multi-target collaboration with Bristol Myers Squibb in 2024, Launched Neo-1 foundation model for molecular interaction design (Mar 2025), Expanded collaboration with Blueprint Medicines (May 2025), Formed $1B+ strategic alliance with Halda Therapeutics (Aug 2025), Rebranded from VantAI to Proxima in January 2026, Closed $80M Seed funding round led by DCVC and Nvidia's NVentures in 2026

Growth metrics

Significant commercial traction through multi-year pharmaceutical collaborations, including a deal with Bristol Myers Squibb valued at up to $674M and a $1B+ strategic alliance with Halda Therapeutics.

Market positioning

AI-native biotechnology leader specializing in induced proximity and protein-protein interaction design.

Geographic focus

North America (HQ in New York) and Europe (via strategic pharmaceutical partnerships).

Patents and IP

Proprietary computational architectures for geometric deep learning and a growing portfolio of chemical matter patents for specific therapeutic leads discovered via the platform.

About Zachary Carpenter

Zachary Carpenter is the Co-Founder and CEO of Proxima (formerly VantAI). He previously served as a Strategic Lead at Roivant Sciences, where he was a key architect in the 'Vant' model of company creation. He holds a PhD in Chemistry from Columbia University and is a recognized expert in induced proximity and the application of geometric deep learning to protein-protein interactions.

Official website: