@CommureOS just raised $70M at a $7B valuation. The business was already trending cash flow positive. The round was mostly for pricing. @tanay_tandon sat down on Care Shift and laid out the playbook. It started as Athelas. Tanay was 18, building an FDA-cleared blood diagnostic device out of his Stanford dorm. Co-founder Deepika was at Stanford Med Center's imaging lab while Tanay was in the AI lab. The company started at the intersection of those two worlds. @ycombinator backed them. Qasar Younis was their YC partner. @Alfred_Lin Lin at @sequoia wrote the first Series C check. @htaneja at @generalcatalyst was an early supporter. They got the device cleared, scaled to millions in revenue. But while deploying in clinics, the same pattern kept showing up: hours lost to documentation, insurance denials, billing. The real bottleneck was admin, not clinical. Athelas merged with Commure in 2023. Tanay became CEO. Now the platform processes tens of billions in claims annually. 85%+ with no human in the loop. What stood out from the conversation: Point solutions have 3 to 4 years left. Tanay's analogy: grammar-checking software died when word processors built it in. The same thing will happen to ambient, RCM, and back office tools. Purpose-built models are overhyped. Foundation models are so good that training your own is like Dropbox leaving AWS. Better to spend the engineering resources on product. Token costs look expensive against traditional software. Cheap against the onshore labor they replace. That's the real calculus. The Rockefeller playbook: aggressive acquirer, horizontal and vertical integration, very little outside dilutive capital. Voice agents are becoming one of their fastest growing product lines, built on top of deep EMR, RCM, and payer hooks. Best advice Tanay didn't listen to: people told him building a hardware medical device at 18 was a bad use of time. Pragmatically, they were right. But doing the hardest thing first convinced the team they could do anything. Full episode of Care Shift is out now.