Company leader
Sri Somasundaram
Co-founder and Co-CEO
Latent
Sri Somasundaram is a co-founder and co-chief executive of Latent, a San Francisco company whose AI helps health system pharmacies get specialty medications approved and to patients faster. He started the company in 2022 with Rishabh Jain, a friend he met in first grade.
Growing Up Around Medicine
Somasundaram's mother is a physician, and he grew up with an appreciation for healthcare. He met Jain in first-grade homeroom, and he has written that "I’m incredibly fortunate to have someone who has seen me and grown with me through every stage of life."
Research at Stanford and Hebbia
At Stanford he worked in an immunology lab studying the molecular binding patterns behind adaptive immunity, and he built deep reinforcement learning robots, with papers at ICML, ICLR, and Nature. He earned a master's degree in artificial intelligence and joined the founding team of Hebbia, an AI company, where he put language models to work for large financial institutions.
Founding Latent
The idea for Latent took shape in his 300-square-foot New York apartment, when he was sick with Covid and Jain, fresh from selling his last business, was coming off a hard run-in with the healthcare system. Looking back on the start, he said, "We were just really passionate and beautifully naive about wanting to build a better world for patients, and what we could do in healthcare."
The two founders began by interviewing chief medical information officers, finance chiefs, and IT chiefs at leading health systems to learn where patients lost time between a prescription and treatment. Latent started with prior authorization for specialty drugs.
Building Latent
Latent later added appeals, drug prioritization, and autonomous agents that work inside the EHR, and it raised an $80 million Series A in March 2026. By September 2026 it was live in 60 health systems, including Cleveland Clinic and WVU Medicine. He sums up its core this way: "We have an agent that sits in the EMR and is capable of answering questions about the patient data."