Luminai raises $38M Series B to automate health system back-office operations — hits 90% on-prem pass rate

Luminai has closed a $38M Series B led by Peak XV Partners (Sequoia's India and Southeast Asia fund), with General Catalyst and Y Combinator participating. The company, founded in summer 2020, automates back-office and administrative workflows for large U.S. health systems and has reached $100M ARR with a team of 30.

The problem

American healthcare still runs on fax machines. Luminai's core pitch is that AI models have reached the reliability threshold, over roughly the last six to twelve months, where end-to-end workflow automation is genuinely viable at scale. The company targets what Dinakaran puts at over $1 trillion in wasted administrative operations — not clinical decision-making, which he explicitly excludes.

A concrete example is referral processing at Cleveland Clinic. Patients arriving from across the world, including the Middle East, send handwritten fax documents to the Clinic's operations team. Luminai deploys a virtual inbox agent that triages each referral by risk, then processes it automatically, replacing a manual review workflow.

“American healthcare still runs on faxes. There's over a trillion dollars just wasted in administrative operations work that today can be done by software systems... We've offered on prem to every single health system we work with. Really? But only 10 to 15% of them have actually taken us up on it.”

Go-to-market

Roughly 20% of the team comes from Palantir, and Luminai uses a forward-deployed model, embedding staff directly inside customer operations to map and automate individual workflows. Dinakaran says that depth is unavoidable given how nuanced healthcare operations are.

On-prem reality

Luminai offers on-premises deployment to every health system it works with, expecting data-sensitivity concerns to drive demand. In practice, only 10–15% of customers take up the option. The on-prem checkbox matters during procurement — health systems want to know it's possible — but most end up running in the cloud anyway. That suggests the commonly assumed regulatory barrier to cloud adoption in healthcare may be softer than it appears, at least for back-office workflows.

Data scale

Healthcare generates roughly eight times more data than the next largest U.S. enterprise industry, and over 90% of it is unstructured — contracts, PDFs, handwritten notes, loose text across legacy IT systems. That unstructured volume is both the core problem Luminai is solving and the structural reason the opportunity is large.

Dinakaran has been building since YC Summer 2020, a five-and-a-half-year grind that positioned the company to absorb the AI capability step-change of the last year. At $100M ARR with 30 people, the revenue-per-head figure is notable even by enterprise software standards.