Latest / StartUp Health NOW Podcast / Meet Cognimark: How Andrew Lu Is Reimagining Brain Health with an AI-Powered Neurology Copilot

Transcript

Meet Cognimark: How Andrew Lu Is Reimagining Brain Health with an AI-Powered Neurology Copilot · StartUp Health NOW Podcast

• 0:00
  [on hold music] Welcome. My name is Unity Stoakes, co-founder at Startup Health. Today, we are speaking with Andrew Lu, the CEO and co-founder of Cognimark.
• 0:15
  They are part of the Alzheimer's moonshot at Startup Health, and we're gonna be talking all about their amazing platform that came out of some research from Boston University, uh, focused really on AI-powered dementia care and their neurology co-pilot for clinicians.
• 0:36
  Stick around and we will be speaking all about Cognimark with Andrew Lu. [upbeat music] Hello, Andrew. Good to see you again. Great to see you. How are you, sir? Not too bad.
• 0:49
  I'm on the East Coast this week and remembering what winter feels like, which is [chuckles] not as fun as I remember. [laughs] Are, are you in Boston? Uh, I'm in New York, actually. Okay. Cool.
• 0:59
  Is it-- So how cold is it now? We're post-Halloween, pre-Thanksgiving, so this is when the leaves start to change and- Yeah... it starts to get cold. It's not that bad yet.
• 1:09
  It's, I think, mid-40s or upper 40s, so not that bad. But after being in SF for a few months now- Exactly.
• 1:17
  I'm sitting over here in, in Berkeley, and where the sun just always shines, except in the morning when it's foggy. But- [laughs]... I am so excited to be chatting with you today and chat all about Cognimark.
• 1:29
  Just to kind of level set, will you tell us a bit about you and your background? Yeah. Sure. So I'm Andrew, obviously. I'm one of the founders and CEO of Cognimark. Let's see.
• 1:45
  I start with my parents' fun facts 'cause they're more fun than me. My family's originally from China. They went to Taiwan, and then they actually grew up in Brazil, so most of my family grew up in Brazil.
• 1:56
  They came over to the US for grad school in assorted pieces, and have been a family of entrepreneurs.
• 2:02
  My grandparents, my parents all have founded companies and built, whether in Brazil, whether in the US, across everything, telecom, banking, stuff like that. But my parents are both in the healthcare sector now.
• 2:14
  My dad is a physician. My mom was on a board of a hospital in Boston for a while. And I didn't catch the healthcare bug until after college. So I did my undergrad and grad at MIT.
• 2:27
  I studied economics primarily, but also some computer science and some finance, and I was really interested in how businesses can be leveraged to make a lot of outsized impact in the world.
• 2:39
  So I was actually an economist by training, did a bunch of research with some Nobel laureates, was focused in the ESG world, and that's where I thought that I belonged and wanted to build in.
• 2:51
  Went to McKinsey initially thinking of ESG work. Unfortunately realized that ESG and industry is a bit of bullshit, if you'll excuse my [laughs] French, and was disillusioned with that.
• 3:03
  But started doing more healthcare work and found that it was much more impactful and meaningful for myself.
• 3:09
  Started specializing in Medicare and Medicare Advantage, so started seeing a lot more about the eldercare journey and the opportunities and challenges there.
• 3:17
  And what year was this approximately when you graduated MIT from grad school and then- Yeah... went into the, the healthcare track? 20, 2020, I graduated undergrad. Within the year, started doing healthcare.
• 3:30
  Was at McKinsey for three years, uh, specialized in Medicare Advantage. At the same time, my grandmother has late stage Alzheimer's. She's actually entered hospice in the last few days, unfortunately.
• 3:42
  Um, but through the work- I'm sorry...
• 3:44
  and through-- Thank you, a- and through kind of the family journey and we have two doctors in the family, still caught it late, still were scrambling to figure out what to do about it, whether a diagnosis, whether by clinical trials and treatment.
• 3:58
  And so the combination of those two pieces really got me excited about dementia, Alzheimer's, trying to do something about it.
• 4:06
  Went back to do my MBA at Sloan with the express purpose of founding within healthcare and ideally within eldercare, and then was very fortunate to meet my co-founder, Professor Vijaya Kolachalam.
• 4:18
  He's a professor of medicine and computer science at BU, and his lab had been developing our core technology, which is this proprietary AI system that can intake any modality of data, so EHR data, blood data.
• 4:34
  It can take in neurocognitive test imaging data.
• 4:37
  And the idea is really any kind of data, in the future wearables, voice data, and give you these detailed neurology reports that can really arm a clinician to actually have impact.
• 4:48
  And so we started talking two years ago, getting to know each other, launched full time like six months ago. It's just so much to unpack here. Yes.
• 4:58
  So fir- first of all, I went to Boston University, so I'm already a- Oh, nice. [laughs]... a huge fan.
• 5:02
  And I've been a huge fan of your work because, as you described, I think your mission of early detection of people with either Alzheimer's, dementia, is just a critical one. We need to solve this.
• 5:17
  So you've been a part of the Alzheimer's moonshot and related dementias here at Startup Health for, I guess, over a year now. Yeah.
• 5:24
  And now you're officially, as you mentioned six months ago, taking this research that came out of Boston University and all the extraordinary, I think, learnings that you've had.
• 5:36
  You mentioned family of healthcare on both the system side, but also the doctor, clinician side. Um- Yeah...
• 5:45
  but your, your family experience as a, an entrepreneur, but then I'm sure what you saw at McKinsey and just being within the belly of the beast, so to speak, of healthcare was quite- Yeah...
• 5:59
  illuminating because I think a lot ofLet's call entrepreneurs, clinicians, entrepreneurs come in and don't fully understand the business of healthcare. So maybe- Yeah...
• 6:11
  talk a little bit about, about that aspect because you were really coming in eyes wide open to a lot of the business of care, and then you decided to build a startup.
• 6:24
  What did you notice when you were-- before you had set the company up, and as a management consultant, what are some of the things that you noticed even before maybe you had the idea for Cognimark? Yeah.
• 6:37
  Ho- honestly, I'll start like before I, I started or right when I started consulting, my worldview was I'm a data person, went to MIT, like very kind of data-driven, and my naive worldview was people were using data to drive decisions.
• 6:52
  And every kind of decision you have from the CEO of a company to a director to every step is driven by strategical, backed by data, and a kind of thoroughly vetted process.
• 7:04
  And you can imagine my shock when I entered the workforce [chuckles] and especially joined McKinsey, which is often working with companies that are at a struggling point or on, on wanting to change and seeing how kind of decisions actually get made, whether in a good way or in a bad way, the process, and a lot of the challenges that happen.
• 7:23
  And I think what I discovered was the world's a much more organic place than I had thought, probably a good thing, where people say, "I have a problem."
• 7:31
  People come in and say, "I'm gonna offer you, at least in consulting, a, a way to think about it and a way to, to think about fixing it."
• 7:41
  And that was what I spent like a year or two years doing was, okay, stopping at that point. We have a big problem. Let's see. I might not know anything about it.
• 7:50
  I worked for a big insurance company, top three, for their Medicare Advantage appeals team, knew nothing about Medicare Advantage, knew nothing about the appeals process, and then within two weeks, I knew more than the senior VP who was in charge of their stars re- metrics, and I was revamping some of the work that we're looking at there.
• 8:10
  And so that really opened on my eye, my eyes to the fact that world's a lot less structured.
• 8:15
  There are a lot of problems out there that are being talked around and known to be prevalent, but when you actually get to the brass tacks of solving them, the bridge is actually a lot more kind of crossable because you can become an expert by learning, becoming dedicated, and a lot of people are stopping a little too high level.
• 8:35
  And so that's why I moved out of consulting was you're building PowerPoint decks, and you're advising people on how to solve problems, but that's usually where the process stops, and you don't actually get to go and operate and solve and actually execute.
• 8:48
  And so that's one of the big reasons why I started thinking of entrepreneurship was because of that ownership, because of that ability to actually execute and see a solution end to end that was going to deliver value rather than talking around on a big problem, so.
• 9:04
  Yeah. Thank goodness there's people like you in the world that are builders that can translate the amazing research and ideas and innovation, and there's so much extraordinary reshur- research. Absolutely.
• 9:17
  But it needs to be brought out to the real world.
• 9:19
  It needs to be translated out to where everyday people, whether patients, whether clinicians, whether the system itself can get access to it, and it sounds that was a spark that really hit you. I wanna go back to
• 9:34
  something else that you said and not skip past it because I think it's fundamental here, which is you understand Alzheimer's and dementia. You've got this situation in your family.
• 9:45
  A lot of people are dealing with this today, and this is something that touches a lot of people's lives.
• 9:53
  Did you s-- W- did that play into your decision to focus specifically in the area of brain health, dementia, et cetera?
• 10:04
  Because one of the things we see often is many of the best health transformers are driven by purpose. They're driven by passion.
• 10:12
  What was that like for you, and did that play into your decision to kind of focus in this area? Yeah.
• 10:20
  A- absolutely was a huge p- driving force in switching career paths and building in this space and also driving force of what we're doing today, I think.
• 10:29
  When you see a problem and it's much more personal to you, you're much more motivated to get in deep and really explore all the nuances of the problem, especially as you experience it firsthand and see the challenges your family member's going through, the caregivers are going through, and it becomes a much more real problem from something than just on the slide.
• 10:50
  So I think that was a big motivator for my transition to the work that we're doing, and then I think entrepreneurship is hard. Yeah. There's so many ebbs and flows.
• 11:01
  Some days we have a really great conversation or land a contract, and it feels like the top of the world, and some days it feels like nothing's going right, and I think that's a natural process of building. And
• 11:14
  the key is how do you separate signal from noise?
• 11:17
  And in a noisy world, you need some strong internal signals to keep you going, even when, yeah, this month or two is a challenge and we haven't made the outcomes that we wanted, we still have the internal drive.
• 11:29
  We still have the long-term vision.
• 11:30
  I think it's incredibly powerful and necessary kind of fuel to the fire to, to continue building and pushing through, especially when you know that there is value in what you're doing, and it's just a matter of convincing other people, going through that process, and building out and proving it out, so.
• 11:48
  Let's see if we can get the word out here. Tell us, and you started to talk about this, but so cognimark.ai, right? Yeah. You're really this neurology co-pilot for clinicians. Can you unpack that for us?
• 12:03
  Tell usWhat you're building and how you see this really either helping really with early detection in primary care, I'm, I'm supposing, but you're really also building at a time that's part of this AI revolution.
• 12:20
  I know that's p- a big part of your solution, is really an AI-powered dementia care that you're en- envisioning. So can you walk us through the vision and describe really what you're building for clinicians in the world?
• 12:36
  Yeah. Maybe I'll start with kind of our view on the linchpins of the problem- Yeah... and then segment back into what we're building.
• 12:43
  But everything we build comes back to the idea of having outsized impact using business or research, and we think that dementia, Alzheimer's, and ma- a lot of neurology conditions really face a few issues.
• 12:56
  The care journey's broken and fragmented because the care journey starts too late by relying on patients and their caregivers. So for my family, it wasn't until we suspected issues that we brought it up to PCPs.
• 13:09
  Even with PCs, PCPs in the family, that can be a years-long process. For primary care physicians, yep. Exactly.
• 13:16
  And then the second part is primary care physicians are dealing with so much, they don't have the time or the bandwidth to learn about the best practices in dementia, let alone implement them.
• 13:26
  When you're dealing with so many CPT codes, all the documentation that's needed, all the follow-up that's needed, there's a lot of manual work and knowledge that w- should be distilled and brought to the fingertips of clinicians.
• 13:40
  But in today's world, my dad's working 12, 16 hours just on his regular patient load. They don't have time for new tools. They don't have n- new time for other pieces.
• 13:50
  And the last piece is a lot of the tools today are reactionary and don't work well together, so you come across this patient journey that starts too late.
• 13:58
  The first triage point isn't equipped to handle that, and then as a result, all these tools aren't working together to give you a clear and flow-through kind of patient journey.
• 14:10
  And so what Cognimark is building is trying to address those exact linchpins, where we're spinning out of BU and MIT this research that is a proprietary machine learning model that can intake any modality of data and then output these detailed neurology reports.
• 14:25
  That's our key competitive advantage that we're using to start the patient journey really early through risk stratification, plugging base straight into the EHR.
• 14:34
  So I'll walk you through the Rhode Island clinic we're working with. We- Yeah, that'd be great. Get... The more you can make this tangible and where does that data come from? Is it wearables? Yeah. Is it f-
• 14:46
  from a health visit? Is it any- anything you can describe to make this really tangible. Yeah, absolutely. So I'll start with today, and I'll go into the vision of the future. Please.
• 14:56
  And so today with this site in Rhode Island that we're doing, that we're... and several others, is we plug into the EHR.
• 15:03
  We use all the data that's there, whether it's EHR data, we can also plug into imaging, neuropsych tests, and then we use that to feed into our model and generate these reports that tell you the type and stage of dementia patients have and break it down into ways that neurologists think about it.
• 15:21
  And so if I go to the output so you can clearly see, it's okay, MCI, dementia, nor- normal cognition. We give you a percentage score.
• 15:31
  You can click on it, and then we break it down into the factors neurologists look for that are saying, "These are the patient data points that support this. These are the patient data points that don't support this."
• 15:42
  And we can also look at Alzheimer's, Lewy bodies, vascular, all these pieces. That's our core technology on helping the physician evaluate dementia.
• 15:52
  And then what we're doing, this workflow is we risk stratify, find of your hundred thousand patients, these are the five thousand patients you should see as soon as possible.
• 16:02
  We go to the patient visit, and then we provide our neurology reports and say, "Hey, clinician, these are the conditions that they might have. These are the reasons why.
• 16:12
  Here are appropriate gold standard tests that people often use in your position. So here are neurocognitive tests like MoCA, Mini-Mental, MMSE that you can use.
• 16:22
  Here are the new blood-based tests that people are pioneering for rule in and rule light diagnosis. And then here are other steps you can do, like MRI, CSF, PET, et cetera." And then
• 16:35
  after we arm them with those options, and we actually are partnered with many of these solutions to route them to specific options, we can say, "Okay, you have your 30-minute patient visit.
• 16:46
  Here are the extra questions you should ask so that we can help you bill for dementia-specific CPT codes, and then we'll do the paperwork for 99483, which is the one-hour comprehensive dementia care and caregiving and diagnosis plan.
• 17:02
  Can we automate that paperwork so that using our AI neurology assistant, we're giving you a lot of the insights to give you a first draft that you can check over and prove and make it super easy, and then we route you to the appropriate next step, whether therapy, neurologist, clinical trials.
• 17:17
  We're building out partnerships here.
• 17:19
  So it's all getting to this idea of we need an end-to-end solution that's integrated, and Cognimark ideally is going to be the guide that starts the journey early in a way no one else can, plugging into the EHR and re-stratifying and then helping you at each step of the way, routing you to appropriate...
• 17:39
  Presenting you the options and routing you to our partners in those options. So I wanna double tap on a couple things, because you said a lot of really interesting things there. So the actual
• 17:51
  user, you mentioned these neurologists, the clinician is using this in their workflow. It's improving their pro- process. Yeah. But your customer of who this is benefiting is also the system or the payer.
• 18:04
  Is that right?That's completely correct.
• 18:07
  So our users, the people who, you know, we're building Athena Marketplace-- in Athena Marketplace, so you can download the app, is a primary care physician, is a neurologist who can use it to help them bill and improve their workflow.
• 18:21
  We wanna sell eventually to payers and ACOs who see the most economic benefit from risk adjustment, from stars and HEDIS improvements, from long-term cost savings, from being able to intervene early.
• 18:34
  Obviously, they have long sales cycles and need very thorough ROI use cases in order to get to that process.
• 18:41
  So we're right now piloting with several direct PCP sites and clinical trial sites to build those use cases and that ROI in order to pilot with the, the payers.
• 18:52
  But I think that's our current thesis on what is gonna be the most impactful and sustainable way for us to grow quickly.
• 18:59
  We have so many partners, and we're being this kinda guide where some of our partners are offering us incentives and other monetization pathways, such as not a referral, but like kinda marketing spend to route them to certain kinds of testing, to teleneurologists, to other options.
• 19:17
  And from pharma companies, they're interested in building out this ecosystem. They can't build capacity, as Phyllis has said, but they're funding efforts to try to integrate this.
• 19:28
  And that's Phyllis Ferrell, who's the chief innovation officer of the Alzheimer's Moonshot at Startup Health. You're walking through a really important case example that I wanna
• 19:41
  kinda mention because fifteen years ago, this amazing gentleman, Bob Kocher, who also was at McKinsey, and then he worked in the Obama administration, he told Steve and I and our health transformers at Startup Health, he's now at Venrock Ventures, but he pulled out an amazing, very simple worksheet, which I think all health transformers, all innovators in this category just need to memorize and focus on.
• 20:11
  It sounds like you, you have. Which is in healthcare, as what everyone really building in this sector needs to know, is who uses
• 20:20
  a solution isn't always the same as who pays for it and then who benefits from it ultimately.
• 20:29
  They might be, in some cases, all the same user or company or system or even patient, but in others, it's oftentimes completely different.
• 20:41
  And you, I think, did a very illustrative job, and that was very helpful, of sh- describing how Cognimark can really help the primary care physicians, can really help the neurologists, the clinicians through that journey.
• 20:56
  It benefits the patients if there's early detection, right? Absolutely. And
• 21:02
  then you're streamlining that process for them as well, saving them time and billing and CPT codes and all that, which that integration right there, the early detection and the integration, is a huge benefit to the systems as well as the payers.
• 21:22
  So everywhere along that value chain, there's different sort of drivers of value, and I, I just wanted to double tap on that because oftentimes I think a lot of extraordinary products and solutions come to market, but they haven't thought through the way you have really the value chain of who's using, who's benefiting, and who the customer payer really is for that solution.
• 21:50
  So I think that's a fundamental thing here. So walk us through where you are. You're... Are you raising money? Are you looking for partners?
• 21:59
  Wave a magic wand scenario 'cause a lot of amazing investors as well as buyers within the marketplace are gonna be paying attention here to what you're- Yeah... building.
• 22:10
  And a lot of people care specifically about Alzheimer's, related dementias, and the impact that you're, you're making.
• 22:19
  If you're my genie with three wishes, and as a consultant, they always say good things come in threes, I think the pieces that we're thinking of is, one, we are kicking off our fundraising.
• 22:30
  We're doing a three million dollar seed round in order to really support a lot of the interest and growth that we're seeing and build out our team to match that pace.
• 22:40
  So introductions or getting to know investors who are passionate about the space and interested in supporting a mission like ours i- is one thing on our wish list.
• 22:51
  The second is we are building, like, a partnership community, and StartUp Health has been so amazing in helping us in this mission as well, and many of our partners are through StartUp Health.
• 23:02
  But as I've been talking about, there is this need for an end-to-end solution. Like, adoption from PCPs, adoption from the system is not gonna happen without this unified end-to-end solution.
• 23:15
  And one of my mentors at a payer system said, "I get five hundred emails from startups every day, and we don't see many of them," because the fact is there's so much going on, it's so much inertia and friction to get things moving, that even for something like neurology, like dementia, which is a huge kind of group in and of itself, they need a all-in-one solution for that.
• 23:40
  They don't wanna deal with seventy vendors and seventy different un- unconnected solutions.
• 23:45
  So we're looking for partners to help build that ecosystem, whether it's blood-based biomarkers, whether it is neurocognitive tests, whether it's teleneurologist and therapeutics.
• 23:57
  We are going to try to be the guide, and that's where we see ourselves being the most helpful of starting the journey and being the guide.
• 24:04
  But we're gonna need our partners to develop this ecosystem to be able to deliver on that end to end.
• 24:10
  So I think that's our second wish list is any partnership exposure that people want to work with on any of these angles.
• 24:16
  We have several and are always looking to expand more.And then the third piece is really on customers and getting adoption going. Yeah.
• 24:24
  Um, and so we've been talking with DAC, CEOI, a lot of these groups that are doing some amazing work on like doing the guide model, showing how it works, and building it out in a manual way to prove to the government, to prove all the stakeholders that this should be done and can be done, and here are the benefits.
• 24:41
  And we're basically like AI-ifying the brain health navigator that DAC is building.
• 24:47
  But we'd love for partners to help us who are interested in really being on the forefront of dementia care and innovation, of testing the product, iterating, and really building with us, because all of our ideas, all of the functions that we build come from the customers and come from the simple question of, how can we be more helpful to you to help others?
• 25:09
  And so any partners a-along the commercialization or the testing is our third bucket of ask. And they should go to cognimark.ai. You're located where? You're on the East Coast.
• 25:23
  We're headquartered in Boston, but working in Boston and SF. Okay. And every once in a while I get you out here in, in the Bay Area, and I know you're back in, in Boston a lot as well, and you're in New York now.
• 25:36
  Can you just quickly describe to close out here, you've been such an amazing contributor to the Alzheimer's Moonshot community. What's it like?
• 25:45
  You mentioned being an entrepreneur is challenging, being a health transformer is, uh, one of the most challenging sectors to build in. You guys are passionate.
• 25:58
  You're backed by science and extraordinary research that's come out of Boston University. You came out of MIT.
• 26:06
  What is it like to be part of a collaborative community that all cares about that North Star of Alzheimer's related dementias? What's that been like from your experience?
• 26:19
  And maybe just share some insights with fellow listeners here. Yeah.
• 26:24
  I think-- So my, my favorite parts about MIT, about McKinsey, were always the people and the cl- community, and being a part of a community that shares similar values, that kind of looks at problems in maybe different ways, but with the same kind of mindset and goals, is so incredibly valuable, especially a collaborative mindset, which is thinking of, this is a huge issue, we're gonna need all hands on deck, and the ocean rises for everyone together.
• 26:53
  And I think that's exactly the community in the Startup Health Alzheimer's Moonshot program. I'm sure in the other Moonshot programs, as we've explored partnerships with, for example, the diabetes Moonshot program.
• 27:05
  But I think it's so incredible to, as a founder and entrepreneur, have a network of people who are going through the same journey, who are also passionate about dementia and Alzheimer's. I always say it with my friends.
• 27:17
  I love talking about dementia and Alzheimer's with-- It's not always seen as a fun topic or an- Yeah...
• 27:22
  intriguing topic to talk about, but having a community of people who are going through the same process, who are having the same goal and willing to listen in, spend extra time to give their advice, their perspective, is invaluable and really helps smoothen the journey.
• 27:38
  And then even on the professional setting, like so many of our partnerships, our commercialization opportunities have come from the community as well, and that's just a natural part of everyone working together to get to the same problem.
• 27:51
  We're very thankful to the Startup Health team, to the Gates Ventures team that, that helped advocate for us to join.
• 27:58
  In joining the community and going through it, our progress would not have been as smooth or as far without the help of the community. Well, Andrew, thank you for everything you and your partners are doing.
• 28:09
  Love to your grandmother and to your family and to all the families that are really dealing with Alzheimer's dementias.
• 28:18
  It's a really cha- challenging issue, as we know, and it's people like you, Andrew, and your team and all across the Alzheimer's Moonshot community and anyone really working in this field, all the amazing researchers, clinicians, innovators that are really pushing progress forward.
• 28:37
  So I thank you personally and from everyone at Startup Health, thank you very much. Andrew Lu, Cognimark CEO and founder. Please support them. Check them out at cognimark.ai. Thank you.
• 28:53
  [outro jingle] And finally, an invitation to all of the extraordinary health transformers, whether you are a startup, someone building the future of health, or whether you are an investor, funder, or buyer of the solutions out there in the marketplace, this is an invitation to you to join the Startup Health member community.
• 29:16
  Learn more at startuphealth.com. Look forward to hopefully seeing you soon. [outro jingle]