Building a company before all the pieces are in place forces calls on timing, focus, and trust. Concurrence co-founder and CEO Ali Khokhar joined Madrona Partner Sabrina Albert on Founded & Funded to share how he has made those calls while building toward what he calls superintelligent care.
Concurrence is building a continuously learning system that understands what drives patient outcomes and helps providers deliver that care, with the goal of health systems that learn from every act of care they deliver.
What founders can take from this conversation:
- How to test whether you’re early or wrong.
- When “go narrow and deep” stops working.
- How to say no to obvious money. Friends pointed him to revenue cycle management because the ROI is clear. Ali passed and says the opportunity costs are already paying dividends.
- How to sell to skeptical buyers.
- How to collect net new context.
Founders building ahead of the market, in a regulated industry, or for cautious buyers will find specific decisions on timing, focus, and trust that they can test against their own company.
Listen on Spotify, Apple, and Amazon | Watch on YouTube.
This transcript was automatically generated and edited for clarity.
Sabrina: Thanks for being here, Ali.
Ali: Thank you. Excited to be here.
Sabrina: In an era of unprecedented innovation, plenty of founders are building for capabilities that don’t yet exist. You’re doing it in the health and medical sector where the stakes are huge, and historically the buyers have been slower. So despite all the tech and data that’s out there, healthcare companies are not learning from the patients they treat. I think this is a problem that would sound pretty straightforward to solve, but it’s been a very persistent problem over the last many decades. I think Concurrence is perfectly situated to solve all of this. So maybe we start by understanding what it is that you’re building today that wasn’t possible before, and why are you sure it will be?
Ali: The way I think about it at Concurrence is we’re trying to build a causal understanding of human health. Historically, this has never been possible, especially because inside of a patient record or a patient summary, all it is is really summarizing what that visit was, especially because it’s very anchored around billing and compliance. It’s not really anchored around driving patient outcomes. Now, for the first time, you have this new technology that is actually able to sit in the line of care delivery and capture a lot more richer understanding and context about a patient.
That really rich understanding then allows you, for the first time, to get a good sense of what is actually driving certain outcomes. And so because of this sort of core technical breakthrough for the first time, we’re able to capture net new information, make sense of it in a way that was never possible before, and as a result, sort of understand how to drive certain outcomes.
Superintelligent Care vs. Medical Superintelligence
Sabrina: And I’ve heard in the market this term called medical superintelligence. Is that a different way than you’re thinking about it, this idea of causal understanding? Maybe unpack that a little bit more for us.
Ali: When most people think of this term, medical superintelligence, which is very loaded, they usually think of this monolithic system that is superhuman and better than any doctor, any provider. We don’t actually necessarily think of it that way. We think of it as more of a continuously learning system that is really focused on understanding what is driving certain outcomes and then being able to actually deliver that care in a meaningful way. And in some cases that might mean actually helping those providers deliver that superhuman care. We actually prefer the term super intelligent care versus medical superintelligence because it’s more around the system that is continuously learning, continuously evolving, where you have a really rich understanding of every single patient to then actually deliver that end care.
Sabrina: Okay, so taking a step back, what does this actually mean for health systems and providers that you’re working with, and how can they actually deliver tangible care?
Ali: So what we usually do when we start working with some of these larger health systems is one, build this unified understanding of their entire ecosystem with a very patient-centric view. We call this the health world model. And then on top of that, we work with them to build very bespoke, very custom patient-facing or provider-facing agentic workflows. So you can think a patient that’s coming in to a visit. Before a visit, there’s a bunch of things that need to happen. They need to make sure their appointment is booked. We do intake, triage, and so on and so forth. The visit itself is a really rich, important piece of context that we want to capture and support and give a lot of information to that provider.
And then after that visit, everything that takes place from making sure that they’re actually going, if they were referred to a specialist, that they actually go have that appointment booked, they can actually show up if they’re prescribed medication. Side effect management is a big piece. Medication adherence is a big piece. So a lot of our focused solutions end up being across that patient journey and sitting in the line of care delivery, but very often supporting that health system across the board.
How to Tell Whether a Startup Is Early or Wrong
Sabrina: So we talked about your framing of this idea of super intelligent care, and when the conditions are finally right, I think they unlock a series of breakthroughs that can make an idea seem almost inevitable. And I’m curious, what is the founder’s job in all of this — how do you distinguish being early versus being wrong, and what would you watch out for?
Ali: As a founder, the thing I think about always is that being too early is essentially the same as being wrong. And so we started thinking through what the right sort of conditions in this space right now are that allow us to have the belief that this end state is going to be inevitable, the end of reality, and that we should start building towards it. But at the same time then, what are we doing today tangibly to continuously build towards that future? And so at the macro level, you can think of it as a massive industry mindset shift around adopting net new technology. That’s for sure the case in healthcare. The regulatory environment is changing. The FDA is way more open and comfortable with a lot of these AI systems being and sitting at the line of care, helping drive patient outcomes. There are obviously technology breakthroughs. These foundational models are continuously getting better and better.
They still need the right system and orchestration around them, and that’s where we focus a lot of our energy on. But most importantly, there’s this thought around, okay, can we actually deliver value today to that end patient, to that health system? And that’s very clearly yes, through a lot of the use cases, like I mentioned, across that patient journey while also having and building the foundation towards this end state future that is almost inevitable. And so we sort of think of it as making sure that there’s short-term value being delivered and the product that we’re building is very intentionally built so that that super intelligent care world is inevitable for us.
Why Go-Narrow Startup Advice Can Hurt In Healthcare
Sabrina: And as you situate yourself for this future playing field, we talked a little bit earlier about avoiding small poisonous ponds. What does that mean when you put it into practice?
Ali: Yeah, I think a lot of founders, especially in healthcare, fall into this trap. Maybe I’ll help define it a little bit. The way I think about it is there’s a lot of very narrow ponds we think of them in the context of healthcare where folks are building companies. And so a lot of my founder friends have said, “Okay, let’s pick one sort of very specific industry, one specific vertical, orthopedics, dental, behavioral health, whatever the case may be. Let’s then pick one specific workflow and just build a product and company around automating that.” Now, that is actually standard conventional startup advice: go super narrow and super deep on a specific problem. I think the challenge for why it does not work in healthcare is that as soon as you go into that, you’ll start realizing that to actually drive end patient outcomes and deliver a lot of value, you need to touch many different problems across that patient journey, across even many different industries, right?
Because you think of it as a patient, you’re not actually saying, “Oh cool, I want to go and do therapy at this provider specifically.” Then you’re also going to Emerge, and then you’re also going to your PCP, and that same patient is having various touchpoints with the healthcare industry. Companies who are then saying, “I’m just going to automate this,” end up getting stuck because you then have all these other problems in terms of retaining context, supporting that patient across various aspects of that journey inside of that ecosystem, inside of that health organization, but then also beyond. And so we then say, okay, building in that very narrow way, we think of it as sort of this small poisonous pond, almost where you go in there with the right initial mindset around conventional startup wisdom, but then you get caught.
And so we took the approach of if we just zoom out and look at the entire industry as a whole, what are the foundational building blocks we should be building from an infrastructure standpoint, from a safety/reliability standpoint, from an understanding a patient across their entire life standpoint, that then can really, really drive lots of value and impact. And so that’s one thing that we always think about is how do we make sure we’re not falling into that trap?
Sabrina: Okay. So tell us about a pond that you deliberately walked past. It’s obviously hard sometimes in the moment to walk past these things when you think that there’s a really big opportunity ahead of you, but as a founder, it’s your job to think about prioritization of these things. What did it cost you, and how do you think about the trade-offs?
Ali: Oh my God, a lot. We’ve had to make some very tough decisions around how we’ve defined our product and platform strategy, and which opportunities we’ve said no to. An obvious example is RCM, right? Everybody in the industry, all of my founder friends, often just say, “Just go to RCM, that’s where a lot of the money’s at. Just go focus your energy on that. The ROI is clear and so on and so forth.” And that’s fair, and that’s true. However, we wanted to really, really focus a lot of energy on what it actually means to deliver care. What does it actually mean to drive those outcomes?
And being and sitting at the point of rev cycle management or prior auth or even ambient listening, all of these sort of point solutions make sense and can be very lucrative in the short term, but in the long term can cause a lot of challenges around how you think about then your product strategy, how you think about what you’re trying to build, who you’re selling to, and so on and so forth. And so it’s a lot of opportunity costs that we say no to early on, but it is already paying us quite a bit of dividends.
Earning Trust From Skeptical Buyers
Sabrina: Yeah. And as I think about it, I think one of the reasons people do focus on these wedges, as you were describing it, is because healthcare as an industry has historically been more skeptical by design, and they want to see the ROI, the product right away. But what are you learning as you’re building at this infrastructure layer of why customers are saying yes to you? And I know you guys just announced a big partnership with Databricks as well. So maybe share a little bit about how you make it easy for the customers in a historically skeptical world to say yes.
Ali: Yeah. I always say customers in this space should be skeptical. Skepticism is good. There’s a lot of players in this space saying a lot of different things. And so we actually love working with folks who are quite skeptical. The way we think about it is that trust has to be quite deeply earned in the space, especially in the context of when you’re ever engaging with patients in any way, shape or form. And so we think of that whether it’s in the context of safety or reliability or how we train all of our systems, how all of our customers are part of that equation so they get to really shape it and really understand how these systems are going to behave. That helps us overcome some of that skepticism. I think in the context of what we’re building, having an understanding of a true longitudinal patient record is really important.
That means having access to data that is super sensitive; that means talking and spending a lot of time with compliance and legal teams, and that’s all great. We should be doing that. And so, in an effort to try to make some of those things really easy, we then say, “You know what? We’ll come to you wherever your entire ecosystem is built, whether you’re on a Databricks, whether you’re on an Azure or Snowflake, whatever the case may be, we’ll come in and work inside of your existing governed environments and really help accelerate the journey that you’ve taken already by partnering with those players versus having you go on a net new journey in a completely different direction.” That really helped us overcome the second batch of skepticism around, “Do I have to completely redo my entire business from scratch in order to work with you?” And the answer is no, of course not.
Sabrina: And these companies have already invested so much in building out their platform of choice, be that with Databricks and the cloud partners, and they don’t want to make that transition out. They don’t want to pull out their data. So if you can be at that layer in which you’re partnering with them, it allows for you to have an opportunity with them. As a patient, I’m familiar with my Epic system, MyChart, et cetera. And healthcare as a segment has historically generated more data than a lot of industries out there, but it hasn’t made care better. And I’m curious, why do you think that is, and what does AI actually make possible for the first time?
Ali: Yeah, I think the way I think about it is it’s twofold. One is we’re not collecting enough understanding and information, right? As I mentioned, we’re seeing a patient go in and see their PCP maybe once or twice a year, and really that’s a point-in-time snapshot of understanding that we now have captured. And so, AI, for the first time, allows us to sit at the line of care and actually capture way more information, and the right type of information, at the same time, right? So let’s go very specific. You’re a patient, you came in, you got some medication, and you went home. And now we don’t actually know if this medication is working or not.
And if it’s not working, is it just because it’s the wrong medication, it’s the wrong dosage, or is it because something happened in your life that you stopped taking it altogether? All of that really rich understanding is completely missing. We need that understanding to actually build this causal understanding of human health. And so for the first time now, these systems, if built safely and reliably in the way that we’ve been focusing a lot of our efforts on, can actually safely sit at the line of care, communicate with you, follow up with you, understand why or why not you may or may not be taking some medication in this example, and then connect the dots back inside of that existing system in Epic or whatever the case may be so that that provider has a holistic full understanding of what to do next in the context of your care.
And so it’s both a matter of we’re not collecting enough rich context and data about that patient, nor is it the right type. Once all of a sudden you have this technology, you can do both of those. And then you’re now collecting way more understanding and information than ever before. Luckily, these foundation models, actually, if you think of them as a search primitive, are phenomenal at looking at a really large amount of data and finding causal patterns and starting to think through, okay, what is actually driving certain outcomes? And so it sort of ends up being twofold in that way.
Sabrina: Can you walk through an example of one of your customers and how they actually leverage Concurrence today, and what that actually looks like, and how you’re actually getting a lot of that data because you’re powering a lot of these conversations, as you talked about before, be that in the pre-visit workflows, but often in the traditional visit workflow? So actually having a conversation with the system itself and then post-visit, how do you do the right adherence, follow-ups, et cetera?
Ali: Yeah, happy to do so. I’ll give you one concrete example. We have an organization that’s fairly global. We’re deployed in seven different countries, and we’re specifically in the workflow of post-prescription of a medication, doing side effect management and health coaching. And so in that case, we’ve done 10 million-plus conversations or patient encounters there, and we’ve found that we’ve actually, through our systems, been able to drive up medication adherence almost 3X in scenarios where those patients are actively engaging.
And when that patient then, when we do escalate, when there is some urgent situation that shouldn’t be handled by our systems, that provider now has way richer understanding and context of, “Hey, here’s actually what’s been going on with Sabrina over the last few weeks. Here’s why she needs to come in for a visit. We’ve booked her in, but here’s a really rich context of what’s been going on in her life.” And so that becomes really, really important. And then we went back and did a study with them to understand how are we causing a lot of really good understanding of patients?
We build these digital twins, and we found that almost 60% of the durable knowledge and information and context about each patient was learned not from their devices, their biometric markers, their blood tests, their visits, but actually just through the conversations that we were having. And so it really proved out to us that when you can put these agents safely, reliably in the line of care, you can collect net new context that then is so durable and helps inform the care of that individual. And we can share a lot of that information with that entire care team, whether it’s their PCP, whether it’s a specialist they’re being referred to, whether it’s the nurse or a care coordinator who’s calling them after the fact or whatever the case may be.
Sabrina: And over time then the patient’s experience gets much better because it’s more bespoke to them, the different individuals, and you’re able to extract a lot of that data from existing systems, but then you’re also able to pair it with this model that’s improving over time.
Ali: Precisely, exactly.
Why Amigo Became Concurrence: Expanding a mission
Sabrina: That’s awesome. So when we first met, I guess it’s been over a year now, the company was actually called Amigo, and you guys recently expanded in a lot of ways, and the company is now called Concurrence. I’m curious, what was the thesis behind doing that? The name is obviously the visible part, but what else has changed?
Ali: Yeah, the name was actually the very last thing to change. Over the last year, we’ve really been reorienting the company around our expanded mission. That means the type of talent and people that we hired. So we brought over Jay Shah, who was the chief of physician staff at Stanford Health before to be our chief medical officer. We brought over Hythem Sidky, who was previously a lead and director at the NIH to come in and lead a lot of our deployment function. And so the way we’ve been thinking about the talent and the people at this company has been very much a big shift. And then also I would say the types of customers that we work with and serve.
So we’ve had some really, really amazing conversations and paths ongoing with some large governments abroad, large health systems — one of the largest pharma companies in the world — who all want to go and influence and impact how care is delivered. And so that’s sort of also been another really exciting piece that we’ll share more about soon, but the name is the very last part. Now, what doesn’t change is really a lot of our focus and emphasis that has been since we first met around safety evaluation, what are the rigorous testing we do before any of these systems are allowed to go out and be deployed into the world? And going back to areas that we don’t necessarily focus on, i.e., just helping monetization, but rather focused very heavily on patient-facing care delivery.
What Healthcare Looks Like in Five Years
Sabrina: Yeah, I think safety, patient trust in this vertical, this industry is one of the most important things and cannot be undermined in any way. So I love that about your guys’ focus and the thesis has always resonated with me. And so as we think five years out, what is different for patients and providers and what do you think will have to happen for that to be true?
Ali: On the health system side more broadly, we’re going to see health systems that are continuously learning from every act of care that they’re delivering. They’re able to get way more understanding of what’s happening inside of their ecosystem and be able to really, really deliver good care at a lower cost of that care. On the provider side specifically, we’re going to be able to equip them for the first time ever a really holistic causal understanding of what is happening inside of a patient’s life, what is driving certain outcomes and how they can deliver the best care to that end patient. And then for that patient, the dream of end state precision care is finally going to be real. Really, the feeling of hundreds of providers who are constantly sitting there making sure that they know everything about you to give you the best care possible is a dream that will finally be true.
Sabrina: I’m excited for that future state. I think with the tech and what you guys are building at Concurrence, we can actually get there, which is a big reason that Madrona is excited to be an investor in you guys from day one. And as a patient, I’m hopeful for that experience as well. So thanks very much, Ali, for being here with us today. This was a lot of fun.
Ali: Of course. Thank you for having me.
Related Insights
