Abridge — From AI Scribe to the Operating System for Medicine

In this new Innovation & Startups series, we break down Abridge — one of the fastest-growing AI startups in healthcare today.

📧 Connect with Host
 Host Deepti Kalghatgi : https://www.linkedin.com/in/deepti-kalghatgi/
🌐 Visit: https://idaahub.com

We cover the real founding story (it started with a cardiologist, a patient, and one sentence about dignity), how Abridge grew faster than long-standing incumbents like Nuance, what just happened with Abridge's Nvidia and Eli Lilly partnerships, and why this isn't really a scribe company anymore — it's trying to become the operating system for how care gets delivered and paid for.

We close with three takeaways every startup founder can learn from Abridge's playbook: why being first to market matters less than being first to solve the real problem, why credible partners can matter more than features, and why trust has to come before product in healthcare AI.

🎙️ Topics covered:

• The real Abridge origin story
• Why Abridge grew faster than Nuance
• Abridge's Nvidia + Eli Lilly partnership news
• 3 startup lessons every founder should know

📧 Connect with IDAAHub
 Follow us on:
 LinkedIn: https://www.linkedin.com/company/idaahub/
https://www.youtube.com/@IDAAHUB
https://open.spotify.com/show/3V8Vuhqkibej5fUwtwkUMx
https://podcasts.apple.com/us/podcast/the-idaa-hub-podcast-ai-in-finance-healthcare/id1848710327

Transcript

SPEAKER_00 0:09

Hi everyone, welcome to the Idaha podcast, AI and Healthcare. Today I'm gonna start a new series called Innovation and Startups. This is basically because in this AI era, it's very important for startups in healthcare to know how they can grow with the growing competition and what they can learn from ones who are already making a mark in the market. So we're gonna kick off this session with the biggest innovation, I believe, in healthcare in the recent times is the AI scribe. Let me explain why. So think about it. And not actually talk to you more. So today, physicians get to actually listen, understand the real problem, and on from their aspect, it's after years of hearing about physician burnout, this has finally given them some of their time back. So one startup in this space has caught my eye more than any other, and that's A-Bridge. I believe it's one of the fastest growing startups in healthcare, in healthcare AI today, and the story behind it is worth understanding more and also learn more from A-Bridge. Let's start with the origin story here. It started with Dr. Shiv Rao, a practicing cardiologist at UPMC's Heart and Vascular Institute in Pittsburgh. So one day he saw a patient who left her visit clearly very unhappy. So he asked her what happened. And what he learned was it was not actually about him. What there was nothing wrong with what the doctor taught. It was about um her husband who used to accompany her with every visit, used to take notes with the doctor, was not there that day. So this this is this conversation and what she told actually doctor made Dr. Rao realize how important uh it was to remember the conversations between doctors and the patients. So as Dr. Rao dug more into that, he realized it wasn't really about just missing notes, also. So the impact it was having was much bigger. She told him that, or he felt that the whole hospital system had been designed to strip her and other patients of their dignity. So that sentence alone prompted Dr. Rao to start a bridge. So think about it. These are the most important conversations we have with our doctor and the patient. And imagine um losing memory of these things or having no recollection of what the doctor told us, or from the other side, from the doctor's aspect as well. So in 2018, he partnered with two co-founders, Sandeep Kunam, a Carnegie Mellon robotics graduate, and Florian Metz, a Carnegie Mellon speech recognition researcher, and built a company out of the University of Pittsburgh in partnership with UPMC and Carnegie Mellon. This isn't the Silicon Valley garage story, but it has a roots in a hospital, and that's a very important thing for a healthcare startup. When Dr. Rao pitched the idea to a Union Square ventures in 2019, it was described as SoundCloud plus rap genius for medicine. It was thought to be a very wacky idea, nobody had done it before, but the underlying problem was actually very real. So doctors were spending hours on documentation instead of patient care. So what set ABRIGE apart from day one? If you think about it, the team also built its own clinical research recognition model rather than relying on generic transcription, because off-the-shelf tools couldn't handle real medical conversation. Second, they also introduced a linked evidence, a new feature which would allow a clinician to click at any line in an AI note and trace it back to the exact moment it came from. So by the time Generate Your AI exploded into the mainstream, ABridge wasn't just catching up to the trend, it was already ahead. So based on most recent news, so where is A-Bridge heading now? So if you look into that, from being an AI scribe, it is now heading towards being the operating system for medicine. So how this started and how how is um going to be the operating system for medicine? So let's look back into um what happened on June 11th. On June 11th in New York, Dr. Shev Rao and with other health system executives made the case that ambient AI, which started out as a little more than a transcription tool, was ready to do something far bigger than writing notes. ABRICH announced a strategic investment from Eli Lilly and VIDIA and unveiled what it's called the first AI native clinical clinician intelligence platform. So it's a system that takes the same patient clinician conversation and uses it as a foundation for billing, for clinical decision support, payer adjudication, and in even pharmaceutical trial screening. So isn't that amazing to connect all these things through this one single conversation we have with the uh between the patient and the doctor? And also the it's not just this, it says a lot about um when Nvidia and Eli Lilly are backing up ABRIGE to get there. ABridge already has 300 health systems live on it, including Northwestern Medicine, MRE Healthcare, Johns Hopkins, supporting more than 100 million clinical conversations a year across more than 250 million patients. Alongside Nvidia Foundation model partnership, Dr. Rao also announced a new tie-up with ArtSite. It's an Nvidia-backed smart hospital company giving care teams a continuous feed of room sensor data layered with ambient documentation across an entire hospital state. So isn't that amazing? So the future actually looks very different. So as Dr. Rao put it, we have known all along we wanted to be able to connect the dots across the main stakeholders in healthcare. Because the only thing that matters, I think, in terms of AI's impact on healthcare is business model innovation. And that's a very important thing. A business model innovation, I want to say I want that to sink in, that it's a very important thing for an AI startup to really understand. So, what are the big three takeaways from this news? One, Average is just not trying to get big by adding new features, it's trying to win by changing the con the conversation, it what the conversation itself is used for. As Rao said it, as Dr. Rao said it plainly, the only thing that matters is the business model. That's a much bigger ambition here than just being an AI scribe. Second, Eli Lilly and Nvidia aren't just funding rounds, they are strategic alignment as well. Eli Lilly needs faster trial enrollment, and Nvidia needs real-world clinical data set to train on. And A-Bridge needs both the capital and the credibility. So each side is getting something here. And imagine the conversation between patient and doctor. We can start the clinical trial enrollment right there itself. And the third thing is the same conversation that powers documentation, billing, and trial matching is also some of the most sensitive data in medicine. So A-Bridge is now kind of positioning itself as a neutral infrastructure between providers, payers, and life science companies. And we still have to see whether that trust holds at this scale. But future is definitely looks very interesting and bright for Average. So why this is the most innovative one? And what caught my eye here, as covered already, actually, one of the interesting things is the backing Average received also. Average funding and scale back to scale back this up was roughly 1.1 billion raised and 5.3 billion at 5.3 billion valuation. That's huge for a startup. And well over 100 million clinical conversations a year. And there's real research between behind it. A Brit's own applied science team found off the shelf, off the shelf, speech recognition wasn't accurate enough for clinical conversations. So they actually built their own models, trained purely on medical, uh, because models which were purely trained only on medical literature miss what actually happens in the room. So why does A Bridge stand out among so many AI scribes actually? I would say there are three reasons. One, it's vertically very specific. And here the Nvidia model's only job is understanding the clinical conversation. Being that specific is very important for a startup. Second, to repositioning the conversation itself as an infrastructure here. Instead of the note being the only record, ABRIC actually treats that as the source of truth. It's a source of truth for billing, payer review, and also clinical trial matching. And these are all connecting together here. Third, it's not trying to um automate here, if you see. What it's trying to do is mediate, mediate between all these payer, provider, and life science companies. ABridge is positioning as a neutral ground between provider and payer. So let's start with something like why not nuance, right? I mean, why um it was there even before? Why didn't it get to this level? So there were AI scribes before that as well. Nuance was there, but the real difference isn't nuance ignored medical language. It did not. Nuance's whole business for decades were medical dictation and speech recognition. The difference is what each was built to capture and how it was built to capture. Nuance was built for a clinician dictating directly into the system. A bridge was built from day one for the messier problem, an ambient multi-party conversation between doctor and patient. So we have to understand the difference here. Nuance didn't bring its own generative ambient product to market until 2023, years after A-Bridge had already been building towards exactly that. And think of it from a distribution side, to even grow from a distribution side. Epic didn't just let A-Bridge in, it named A-Bridge as its first pal in the partners and pals program and took an actual equity stake in the company, something it had never done before. Nuance was part of Epic too, if you think of it, but not like this. The level of trust was different, and also the linked trace, linked evidence traceability was another thing that put A Bridge further ahead. So based on what we talked today about A-Bridge, what are the three big lessons for a startup? So, what is it all through? So, here are my few takeaways from what I learned about ABRIGHEGH and how it grew. One, you don't need to be the first in the market, you need to solve the real problem in a way people actually prefer using. So this is this was a real difference because uh dictation versus ambient listening. And also it's that physicians clearly preferred how average had it versus how nuance had it before. So that's where the physician buy-in also matters. Second, your innovation needs a real credible partner who believes in you before they even the market does. Epic's bet on average did more for its trajectory than any single feature ever could. So having the right partner during your growth is extremely important. Third, in healthcare especially, focus on building trust and credibility before you focus on just the product or the solution itself. Because products can be re-replicated or easily built, especially in this AI age. Trust, once earned, is much harder for a competition to take away. So that's it for today's episode of Idahub. If you like this kind of breakdown on startups and innovation, and if this is something you like, please follow. I'm Deepti, host of Idahu Podcast. Please share it with other startups and share your comments as well. Thank you, everyone.