But what was the light bulb moment when you decided, okay, we're going to go for this? Because even in 2018, while academically the runway was, was getting closer, or we were closer to landing the plane of AI, it was still a little bit foggy as to where we were going to land.

Yeah, absolutely.

So we started the company just a few months after Google Research had published their paper, Attention is All You Need.

And transformers were certainly already being used in the research community.

And these transformer-based models that we started with predated LLMs.

And so we started with models like BIRD and BioBIRD, long form of Pegasus T5.

We were spending millions of dollars on expert annotation so that we could fine-tune these models to be able to work for this use case.

High

Tell me about that.

Yeah, I just taught myself Ruby on Rails and built something that could help the residency program.

Other residents and fellows basically keep track of what they were learning on the floors.

So the way residency works or worked for me at Michigan was that you're basically on your own at night.

You're treating incredibly sick patients who are coming to this tertiary care center from other hospitals because they found them to be too complicated.

But you're also you're a trainee.

You don't have years of experience under your belt and you're alone at night.

But I'm just curious, is that a risk you have to mitigate as a bridge continues to grow?

And, and would the message be different based on who the audience was?

Yeah, totally.

The personas are, are, are very different.

So healthcare is a hard sell on the enterprise side.

When you think about healthcare, I would think about health systems as one stakeholder.

Another stakeholder is insurance companies, and then another one is pharmaceutical companies, like the life sciences companies.

If you can find a way to align across all three, like you're transforming things.

... where would you say Abridge is most differentiated or maybe, um, philosophically different moving forward as well?

Uh, how could any generic AI company do that? But our role we can play in that is by giving you a huge foundation to stand on of open models and then start to domain adapt them for these really complex workflows that we're, we're gonna-- uh, you're going to build.

Yeah.

Uh, when we think about workflows, when we think about our product, really- You know, from our standpoint, we just have to deliver the best possible experience.

We have to be able to move all the needles that we were talking about earlier today.

And to an extent, for sure, it means that we are gonna partner with the frontier model companies, and we do.

We have incredible relationships there.

And also, since we have a very, you know, maximalist sort of attitude around sprinkling intelligence everywhere we possibly can, it means we do have to reach down lower into the stack and control our destiny.

And so we've got to be very, very thoughtful about the token economics, and that is, again, giving amazing, um, applications and platforms like yourself the ability to control all of that.

Where do you see this shifting and what, what are some of the great wins you're seeing as well?

Yeah, absolutely.

I think that, um, The technologies have shifted and how we end up solving problems will continue to shift.

But for us at Abridged, the core thesis is that healthcare is human and thus one of the original signals, one of the most upstream signals in healthcare delivery is a conversation.

It's a dialogue.

it's upstream of so many different workflows.

And being a part of that conversation means that we can not only help with clerical work, like documentation, like order entry, like ICD diagnosis, insertion into the medical record, like prior authorization, like clinical trial recruitment, like the list goes on and on and on.

And so as we think through this now, what advice would you give our digital health leaders out in the industry as they're deploying AI and they're thinking about how fast can I scale and the speed in which I can scale, but at the exact same time, Do I do it and lose the human experience? How would you give advice for them as leaders to approach this as we wrap our time up?

And was the model, I mean, Fable 5, I mean, Mythos, do you anticipate these could have major breakthroughs for the clinical setting as well?

Yeah, to an extent.

I think that there are certain tasks in healthcare that these frontier models are just going to absolutely crush.

There's no question about it.

I think the real differentiation for vertical AI companies like ours is getting really, really deep.

into workflows and working with datasets that just no model has ever seen before.

A lot of times, it's not just the data, it's the business logic, and that business logic lives in someone's head in a basement, in the revenue cycle department in a hospital.

Like, how do you think about that trade off? Well, I think, How you wrap the

Doesn't really give me any insight.

Yeah, absolutely.

So just to give you a sense of kind of where things have been and now where things are with Abridge.

It used to be that I'd walk in a room and, you know, before I would walk in that room to talk to that patient, I'd maybe, like, prep the night before.

I'd maybe take some notes to understand who this person is, what are their issues, so that I can feel that much more equipped when I actually walk in.

Then I'd walk in the room, and we'd have all kinds of chitchat.

I think that's what you're referring to.

How do you negotiate, uh, with Epic being that they are such a behemoth in your industry?

space.What did you not expect in GTM that surprised you most?

This is where strong ideas held loosely.

You know, when you first raise capital, maybe you hear an investor give you advice about starting down market and swimming upstream, disrupting over time.

You know, get, get that PMF, find those fast feedback loops.

And it's true to an extent, but you just have to be really careful and mindful because healthcare specifically in the United States, it's not one $5.3 trillion market.

It's a bunch of different markets, and depending on who you're trying to serve, you have to be really careful about how you segment.

If you're trying to serve clinicians, doctors, there's about a million doctors in the country, maybe 800,000 of them are actually practicing.

scalable.Do you have to have an FD process to be successful selling to enterprise with AI tools?

So a- as you've worked so hard to take one of the most important parts away or improve the process of moving from traditional dictation and documentation and note-taking, all this other kind of stuff, to more of this ambient g- generative AI, get this stuff naturally happening, wh- where do you see this shifting and what, what are some of the great wins you're seeing as well?

Yeah, absolutely.

I, I, I think that, um, the technologies have shifted and, you know, how we end up solving problems will continue to shift.

But for us at Abridge, the, the core thesis is that healthcare is human, and thus one of the original signals, one of the most upstream signals in healthcare delivery is a conversation.

It's a dialogue.

It's upstream of so many different workflows, and being a part of that conversation means that we can not only help with clerical work like documentation, like order entry, like, um, ICD diagnosis insertion into the medical record, like prior authorization, like clinical trial recruitment, like the, the, the list goes on and on and on.

But we can also, by being in the conversation, we can bend the trajectory for what happens next.

Mm.

[chuckles]

I remember, I remember that, that conversation too, and, uh, it was like, it was super exciting to also meet another kindred spirit who puts the clinician first as well.

Um-In terms of my journey, we started the company in twenty eighteen, but I'd say it started even before that.

Uh, prior to this, I used to be an executive at a large health system at UPMC, and I got to lead all things related to provider, the clinician-facing innovation.

And that included stuff we were building inside the system, and that also included stuff we were investing into, like companies or even researchers at Carnegie Mellon University.

And that was an incredible opportunity for me over a bunch of years to just learn osmotically from people who could figure out how to do due diligence on a company to invest, but also learn from people who were doing really amazing research when deep learning was really taking off, especially in healthcare.

There was so much excitement in the mid-2010s.

'Cause I'm sure it, it's evolved so much.