Baba raises $40M Series A to scale its human-led healthcare coordination platform for Medicare patients

Baba raises $40M Series A to scale human-led Medicare care coordination

Baba connects older adults and Medicare patients with nurses and social workers — human advocates who handle the coordination work that falls through the cracks of the American healthcare system: benefits enrollment, getting the right appointments, arranging transportation. Connor Sweeney, the company's founder, describes it as a deliberately human service at a moment when that's an unusual pitch.

The $40M Series A was led by FirstMark Capital, with CVS Health Ventures and General Catalyst participating. Sweeney was previously on the show in February for what appears to have been a seed round, making the timeline from seed to Series A less than a year.

“We're announcing Boba's $40,000,000 Series A... we connect older adults and their families, usually folks with Medicare, with other humans — hundreds of experts who are nurses and social workers across the country... we've been able to use a lot of AI but still keep this a very human service... 10,000 patients that you've helped. Have 100,000 on the wait list.”

Scale and demand

Baba has served 10,000 patients and has 100,000 people on a waitlist. The waitlist isn't a capacity problem in the traditional sense. The service is covered by insurance, so the constraint is convincing health plans — Cigna, Humana, and others — to cover it in their markets. As plans come on board, the waitlist clears without Baba needing to manually recruit patients. Early plans that signed on are, in Sweeney's words, "vastly blowing it up."

AI role

Advocates use AI to handle parts of care coordination tasks, but Baba is explicit that full automation isn't the product. The company published what it calls BabaBench — available at baba.com/bench — measuring where current models fail on care coordination. The core finding is that models make consequential errors in healthcare contexts, particularly oversharing information. In a setting where disclosing too much can cause a patient to lose insurance coverage, that failure mode is disqualifying. Human advocates stay in the loop precisely because of these edge cases.

Team

The company has roughly 20 people, around half engineers, based in SoHo, New York. Sweeney says the team has doubled since February and is still hiring for applied AI roles outside San Francisco.

The strategic read is straightforward: Baba's moat isn't the AI, it's the insurance coverage relationships and the trust of families navigating a system most people find impenetrable. If models improve and drive down the cost of the AI-augmented parts of the workflow, that's margin expansion, not disruption.