advicehub
About my work

Serial deep-tech founder and theoretical physicist who builds AI for molecules, materials, and medicines. As co-founder and CSO/CIO of Creyon Bio he built the datasets and machine-learning models that predict the pharmacology of nucleic-acid drugs (ASOs), delivering an N-of-1 medicine from whiteboard to first dosing in under 13 months and co-leading a partnership with Eli Lilly worth up to ~$1B. He now builds polymer and molecular foundation models and glass-box agentic AI systems with collaborators at Stanford and Pitt. He advises teams on AI-for-science and AI-driven drug discovery, nucleic-acid medicine R&D, quantum-chemical and ML featurization of molecules, and how to build and lead high-performing ML and wet-/dry-lab organizations.

Highlights
4
Notable projects
16
Research papers
10
Roles held
Experience

Founder (Stealth-mode Deep Tech)

Current

Vislesy Ventures / Stealth-mode startup2025-10-01 – Present

Physics-informed AI startup building a first-of-its-kind Polymer Foundation Model to design novel, optimal, product-ready polymers (chemical sequences) the way LLM foundation models generate code. Co-founded with Todd Martinez (Stanford), David Pekker (U. Pittsburgh) and Stefano Ermon (Stanford). Also building glass-box agentic reasoning systems for an AI medical assistant, with an initial focus on an AI Physical Therapy Assistant.

Chief Innovation Officer, Cofounder & Board Member

Creyon Bio2024-10-01 – 2025-08-01

Led the interface of wet lab and dry lab as Chief Innovation Officer.

Chief Scientific Officer, Cofounder & Board Member

Creyon Bio2020-01-01 – 2024-10-01

Mission: make gene-centric medicines affordable and rapidly available globally. Pioneered Design-of-Experiments datasets and quantum-chemical featurization to predict ASO/nucleic-acid pharmacology. Delivered an N-of-1 medicine (TNPO2 Foundation) from whiteboard to dosing in under 13 months; partnered with Eli Lilly (up to ~$1B biobucks). Led both wet labs (San Diego and India, plus CRO network) and dry labs (theory, AI/ML, software).
Notable work

TNPO2 N-of-1 medicine

Allele-selective ASO for a single patient, taken from whiteboard to first dosing in under 13 months; well-tolerated and in its fourth year of dosing.

Polymer Foundation Model

First-of-its-kind physics-informed foundation model to generate novel, product-ready polymers (current stealth startup).

Research papers

Polyformer: a generative framework for thermodynamic modeling of polymeric molecules

arXiv:2604.14241

Published: 2026-01-01

NucleoPhi: A Transferable Force Field for Modified Nucleic Acids Enabling Atomistic Simulation-Driven Oligonucleotide Design

ChemRxiv (2026.0116)

Published: 2026-01-01

Toxicity of Antisense Oligonucleotides is Determined by the Synergistic Interplay of Chemical Modifications and Nucleotide Sequences, Not by Either Factor Alone

ChemBioChem 26(20), e202500584

Published: 2025-01-01