Led · Equity · Aug 2026

Lanyon AI develops a neurosymbolic scientific AI that produces implementations and formal proofs simultaneously, claiming outputs that are provably correct by construction. The system uses a large language model to propose formal specifications, then employs reliable symbolic methods to expand those specifications into code and corresponding proofs; if a specification cannot be proven, the code is not generated. The company reports its agent operates at a fraction of the token and compute cost of frontier models while enabling complex physical simulations, theorem proving, and algorithm discovery. Lanyon is initially targeting high-precision, mission-critical domains such as aerospace engineering, space propulsion, and nuclear energy. The founding team includes applied mathematicians and physicists formerly of Princeton and the Princeton Plasma Physics Laboratory, bringing decades of combined expertise. The company emerged from stealth alongside a $10.6 million funding round to advance its technology.

Participated · Series C · Jul 2026

Chai Discovery develops AI models that reason about biological structure and function to generate new molecular designs from scratch and put those models into a software platform used by pharma teams. The company reports partnerships and deployments with major pharmaceutical firms, including Eli Lilly and Pfizer. Its model lineup includes Chai-2, released in 2025 as the first zero-shot generative platform for fully de novo antibody design to achieve double-digit experimental success rates, and Chai-3, which materially improves target success rates and binding affinity versus its predecessor. Founded in 2024 and based in San Francisco, Chai aims to bring engineering precision, speed and scale to pre-clinical discovery. The company recently raised $400M in a Series C at a $3.8B valuation to accelerate R&D, deployment and partnerships.

Led · Equity · Mar 2026

Congruence Therapeutics is a clinical-stage biotechnology company that uses a computationally driven discovery engine, Revenir, to identify allosteric and cryptic pockets and generate novel small-molecule correctors. Its lead program, CGX-926, is advancing toward a Phase 1/1b study for MC4R-deficient genetic obesity, and the recent $39.5M financing is earmarked to support that trial. The company has built a pipeline of wholly owned programs spanning genetic obesity, GBA Parkinson’s Disease, and α1‑Antitrypsin Deficiency. Congruence also leverages its platform to support research collaborations with large pharmaceutical companies across oncology, metabolic disease, neurology, and immunology. The company is led by CEO Dr. Clarissa Desjardins and is based in Montreal.

Participated · Equity · Mar 2026

Earendil Labs applies AI across discovery, design, optimization, and development to generate and progress protein-based biologics programs. Its AI-native platform has produced more than 40 programs and includes HXN-1001, a half-life extended anti-TL1A antibody that the company says is ready for Phase 2 clinical development, with multiple IND submissions planned in 2026 and 2027. The company works through a combination of internal pipeline development and strategic partnerships, including a 2025 worldwide exclusive license agreement and broader strategic collaboration with Sanofi covering multiple programs. Earendil operates alongside affiliate Helixon Therapeutics and targets therapies for autoimmune diseases, cancer, and other areas of unmet medical need. With the announced $787 million financing, the company plans to scale its AI-driven R&D organization, expand scientific and engineering teams, and advance multiple internal programs toward clinical milestones.

Led · Series A · Mar 2026

Tamarind Bio offers an AI platform designed to streamline how life-science teams discover, select, and run computational models. The system currently hosts more than 200 models covering use cases such as protein design and molecular dynamics. Around 100 biotech companies, including pharmaceutical majors Bayer and Boehringer Ingelheim, rely on the software to accelerate R&D workflows. By automating model orchestration, the company aims to shorten experimental cycles and reduce computational overhead for researchers. The newly raised capital will be used to grow the model catalog, deepen AI capabilities, and scale customer support. Tamarind Bio’s business model appears to be usage-based access to its growing library of models, and the firm is positioning itself as critical infrastructure for computational biology. Prior to the Series A, the company had been operating with undisclosed seed financing, and the latest raise brings its known funding to $13.6 million.