This AI Started Taking Notes. Now It's Worth $900 Million.

Two years ago, Heidi Health was a clinical note-taking app. Doctors talked, it transcribed. Useful, not exciting.

This week, the company raised $340 million at a $900 million valuation. That jump tells you something about where AI value actually lives.

From Scribe to Agent

Heidi started by solving a specific, painful problem. Doctors spend roughly two hours on paperwork for every one hour with patients. Heidi's AI Care Partner listens to the visit and writes the clinical note. Simple.

That tool now covers 175 million patient visits across 190 countries in 110 languages. Annual recurring revenue hit $50 million, up from $1 million two years ago.

But the $340 million round isn't about note-taking anymore. CEO Dr. Thomas Kelly is using the money to push Heidi into what the company calls "agentic capabilities." The AI won't just document the visit. It will handle pre-visit prep, pull relevant medical research during the conversation, coordinate follow-ups, and manage administrative tasks around each patient encounter.

A scribe records what happened. An agent does the work. That's the shift.

Why This Matters Beyond Healthcare

If you run a business where knowledge workers spend more time on process than on their actual expertise, this story is your story.

Heidi's growth curve shows a pattern that repeats across industries. Start by capturing the workflow: transcription, documentation, data entry. Then use that captured context to automate the steps around it.

Think about what a law firm's AI could do if it moved past document review into case preparation. Or what an insurance company's AI could handle if it graduated from claims transcription to claims processing. The bigger value comes from handling what happens next.

Heidi figured this out early. Their product suite already includes Evidence (medical research at point of care), Remote (audio capture from patient interactions), and Dictate (voice-to-text). Each piece feeds the next. The more context the system captures, the more work it can take on.

The Numbers That Matter

Blackbird led the $100 million equity portion. General Catalyst's Customer Value Fund put in $240 million as a growth investment. The valuation nearly doubled from the $465 million Series B just eleven months earlier.

ARR growing 50x in two years is the kind of number that makes investors move fast. But the real signal is what drove it. Heidi didn't grow by adding features nobody asked for. They grew because doctors kept coming back after the first visit note. The product solved a real problem well enough that word-of-mouth did most of the selling.

Pranav Singhvi at General Catalyst put it plainly: Heidi "has shown it can win trust" in complex health systems. Trust is the moat.

What to Take From This

If you're building AI solutions or evaluating them for your business, the Heidi playbook is worth studying.

Pick one painful, repetitive task. Solve it completely. Don't try to be an "AI platform" on day one.

Then use the data you capture from that first task to expand into adjacent work. Heidi's clinical notes give them the context to prep the next visit, surface relevant research, and flag follow-ups. Your industry has the same kind of chain.

And earn trust before you scale. Heidi operates in 190 countries not because they launched everywhere at once, but because doctors in one clinic told doctors at the next one. In healthcare, a peer recommendation matters more than any marketing campaign.

The $340 million headline is impressive. The real story is simpler. An AI that takes good notes became an AI that does good work. That transition is open to every industry. Whether you build it or buy it is the part worth deciding now.