Skip to content
YC Root AccessYC Root Access

This Startup Is Automating America's Biggest Hospitals

In this episode of Founder Firesides, YC General Partner Aaron Epstein sat down with Kesava Kirupa Dinakaran, the Founder of Luminai (S20), which just raised a $38M Series B. Luminai is the AI transformation partner for health systems, automating the manual operational workflows for hospitals like Cleveland Clinic that still run on faxes and paper. Apply to Y Combinator: https://www.ycombinator.com/apply Work at a startup: https://www.ycombinator.com/jobs

Aaron EpsteinhostKesava Kirupa Dinakaranguest
Apr 9, 202632mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:00 – 0:05

    Intro

    1. AE

      [upbeat music]

  2. 0:05 – 0:36

    Luminai’s mission: AI transformation for hospital operations

    1. AE

      Today I'm excited to welcome Keshav from Luminai, who is fresh off a Series B of $38 million from Peak XV. Keshav, thank you for joining.

    2. KD

      Well, thank you for having me.

    3. AE

      Uh, maybe to start things off, tell us more about what Luminai does.

    4. KD

      So Luminai is the AI transformation partner for health systems. We help, uh, large hospitals across the US move a large amount of their operational workflows from people to computers.

    5. AE

      Okay. Are there specific workflows, or can you walk us through one of those specific workflows that they're doing?

  3. 0:36 – 2:08

    Case study: automating fax-based referrals and triage at major hospitals

    1. KD

      Absolutely. So, um, so if you think about, uh, someone like the Cleveland Clinic, um, so the Cleveland Clinic is one of the most, uh, prestigious large academic medical centers in the country. And, um, uh, as you can imagine, you know, last year they did a little over 16 million patient encounters, um, where they treated some of the most complex and most nuanced, um, uh, care journeys that, that m- most hospitals in the world don't ever see. Um, and they also have a huge, uh, uh, research arm where they work on some of the most critical diseases that are happening. And as you'd imagine, for someone of that caliber, there are patients from all over the world who want to get care at the Cleveland Clinic. And, um, this is the unfortunate reality about healthcare in America, but, um, the way that most of those patients get referred into, uh, the Cleveland Clinic is through a fax.

    2. AE

      Hmm.

    3. KD

      Um, and, um, if you think about what that means, uh, this is on one end a physician, uh, from outside the Cleveland Clinic, writing a piece of note about a particular really critical, uh, patient and then sending it to the Cleveland Clinic fax line. And then there exists operational teams on Cleveland Clinic, Cleveland Clinic's end, where their entire job is basically to look at these faxes, figure out is this sales spam that's coming to Cleveland, is this like, uh, a thank you note from some random provider, or is this a high critical cancer patient who needs, uh, immediate attention today?

    4. AE

      Hmm.

  4. 2:08 – 3:38

    From unstructured paper to structured workflows: the platform approach

    1. KD

      Um, and if you think about it, uh, that operational process is extremely manual, super heavy. What Luminai does is, for them, we're essentially almost like their frontline, uh, inbox agent.

    2. AE

      Hmm.

    3. KD

      Um, where now every single, um, fax that comes in hits Luminai first. We've become sort of the, uh, the initial triage for all of these faxes, and then if it is a high critical patient, we immediately process them. If it's someone who's not as high critical, but still requires ca- care at the Cleveland Clinic, then we're essentially, you know, extracting all the information, matching it to the right patient and provider within the internal EHR, uh, and then routing it appropriately to the right department, there are thousands of departments within the Cleveland Clinic, and then kicking off sort of the scheduling process. And so this is one example of probably, you know, uh, dozens of workflows across the Cleveland Clinic that, uh, and other health systems across America, where there's... it's all just being done by people, process, and paper. Um, and what Luminai does is become sort of the data transformation layer for these institutions, where we're converting all of this unstructured data, like faxes, into structured data, and then we have a workflow engine on top where you can essentially build a set of verticalized agents to go solve very specialized and very important problems. Um, and through that, hopefully, you know, uh, with the 30% of spend, uh, that's on the administrative waste within healthcare, which is over a trillion dollars, uh, we can get rid of that, so patients actually can get care faster. And then also at the same time, these hospitals can actually run much more effectively.

    4. AE

      It sounds incredibly valuable.

    5. KD

      Yeah.

  5. 3:38 – 4:37

    Growing up in India: finding identity and focus through Rubik’s Cubes

    1. AE

      Um, I, I'd love to back up to the beginning, and I feel like you have such, um, an incredible unique story, and talk about how you got to this point. Like, tell us about the early days growing up in India and, um, uh, I, I feel like, uh, it's interesting to hear what led to some of the motivations and what led you down this path that you're going down now.

    2. KD

      Yeah. So I actually was not, um, uh, very good academically, uh-

    3. AE

      [laughs]

    4. KD

      ... when I, uh, grew up. And, and, um, as you'd imagine in, especially in, um, the Indian education system, if you're not good academically, um, uh, you're sort of part of a separate, uh, group of people. Um, and, um, and so I ended up kind of thinking about, you know, what are the other things that, that seem interesting? And for whatever reason, I ended up coming upon- across a Rubik's Cube. Um, and, um, I ended up basically getting super obsessed, uh, with solving Rubik's Cubes. Um, and, uh, I ended up spending maybe my entire childhood, basically, solving Rubik's Cubes. Uh, and-

  6. 4:37 – 8:44

    Elite practice habits and competitive feedback loops

    1. AE

      How much time were you spending every day?

    2. KD

      You know, um, I got into it pretty competit- competitively, so I ended up break... essentially maybe seven or eight hours every day on top-

    3. AE

      Wow

    4. KD

      ... of, like, middle and high school. So I'd wake up earlier than, three or four hours before school, I'd practice, and then after I'd come back from school, I'd practice, then do whatever final homework I could do and go back to sleep and do it and repeat for, like, seven years-

    5. AE

      Wow

    6. KD

      ... sort of straight. And, uh, as you can imagine, there's some pretty crazy people in the Rubik's Cube community, 'cause they're all quirky nerds. Um, and it was a, it was a great, uh, great group of friends I ended up making. And funnily enough, to tell you how crazy some of these people were, like, I learned Rubik's Cubes from Andrej Karpathy, uh, on YouTube, who eventually-

    7. AE

      [laughs]

    8. KD

      ... ended up founding OpenAI. Um, but, uh, yeah, that's what I did most of my sort of childhood. I ended up breaking a bunch of world records. I was the captain of the International Rubik's Cube Team, and, and, um, and yeah, so decided to spend most of my time sort of doing that.

    9. AE

      That's, that's incredible. And, um, what was, like, your main motivation behind learning the Rubik's Cube and all the time that you, uh, that you put into it? What were you hoping to get out of that?

    10. KD

      I think there was, like, as a 11-year-old, um, there were sort of two... I guess if I reflect now retrospectively, like, there were two things I think that really, um, got me excited to keep solving Rubik's Cubes. I think one was, in India especially-It's a very meritocratic, um, academically meritocratic society in some ways. Um, and whatha- it's changing obviously now. But one of the main things that, that, uh, showed up was, hey, if you're not academically excellent, then you're not really given any attention to. And what was really interesting about the Rubik's cube community is, uh, the first day I walked in, I walked into this room filled with, like, you know, CEOs, and engineers, and doctors, and musicians, and all these people from every walk of life who are all exceptional outside of the Rubik's cube community. Who are all of varying ages. Like, who were... I had friends who were, like, five years old. I had friends who were, like, 65. And, uh, it felt like a level playing field, where what mattered was whether you were good [laughs] at solving Rubik's cubes. And, um, it didn't matter if you were an 11-year-old, uh, and you hadn't, like, done well academically or something. It was just like, "Can you solve Rubik's cube fast?" And when, for some reason, when I k- kicked off, I actually was not terrible at it.

    11. AE

      Mm-hmm.

    12. KD

      And so I'd started realizing I was good at certain types of techniques within Rubik's cubes, that a bunch of the older folks, these are, like, PhDs and, like, uh, doctors, and all these people would come to me and be like, "Hey, what do you think about this particular problem within the Rubik's cube?" And ask me for advice. And I was like, "This is the first time I'm ever, like, experiencing this completely different interaction mode in growing up in the society I grew up in."

    13. AE

      Yeah.

    14. KD

      And so, um, that just sucked me in. Uh, I just felt like, oh man, like, this is the first time where I feel like I have full control over if I'm good at something, then I'm just gonna, like, you know, rise to the top in some ways. So that was the first part. The second part was, it's an insanely competitive, uh, group.

    15. AE

      Mm-hmm.

    16. KD

      Um, and it's incredibly measurable, right? And so, like, f- it was very clear to me. To break the world record, this is the number you had to hit. Uh, to win the competition, this is your average you need to hit. And, like, every day, like, the seven hours I'd practice, it was so clear, like, what my, like, feedback was. Which was like, hey, I landed at this time. Why did it take this much time? What do I need to do to get better? And am I getting better?

    17. AE

      Mm-hmm.

    18. KD

      It was so fast, like, the rate of sort of, uh, that, that, I guess, like, dopamine hit in some ways of every time you solve a Rubik's cube. And, um, yeah, it kept me going for, for, uh, um, almost my entire childhood.

    19. AE

      Yeah. And it's interesting, like, knowing a bit about your background, how you, um, you felt like an outsider. It helped you find this Rubik's cube community. But I feel like it's also put this chip on your shoulder where you're not willing to take no for an answer, and, um, you're incredibly aggressive about going after the things that you want. Talk about, you know, going through the schools and then eventually coming to the US, um, how you were able to pull that off, especially at such a young age. I think you were only 19 when you came to the US, right?

  7. 8:44 – 11:00

    A nontraditional path to the US: UWC, the Silk Road bike trip, and Silicon Valley

    1. KD

      Yeah. So I actually, um, ended up not going to college. And, um, uh, basically what ended up happening was the... Maybe the backstory is, uh, I really wanted to come to college in America.

    2. AE

      Mm.

    3. KD

      And I wanted to do something so crazy that would get the attention of these schools in the US. And so, um, I got a bunch of, like, really, really successful institutions to fund me and, uh, a couple other friends, uh, basically to, um, bike from Turkey to China.

    4. AE

      Mm-hmm.

    5. KD

      Uh, and the high school I went to, which is a separate story of its own, was, uh, the mission of the school was to promote peace. And, um, it was two years. You know, as someone who grew up in India, it was completely sort of out of the blue. I had no idea it even existed. I went on a full ride, mostly because of my Rubik's cubes. [laughs]

    6. AE

      [laughs]

    7. KD

      Um, but, um, ended up basically wanting to live that u- that school's mission. It's called the United World College. And I, um, uh, because I did that sort of cycling trip, um, uh, and what, what ended up happening is a bunch of sort of foundations from other places had taken notice of it, and one of them, uh, had flown me out here, uh, to the US. And, um, my whole dream my, growing up was, like, to figure out a way to come to America. And when I first landed and I came to Silicon Valley, every single thing I thought about, uh, what I thought was true, felt like it was true.

    8. AE

      [laughs]

    9. KD

      Um, and, um, landed in sunny Palo Alto, uh, walked around Stanford campus. And this foundation I went... uh, that funded me had, like, uh, connections to all these, like, pretty crazy, like, founders and executives, and they got me intros to them. I met them. At the end of the 10-day trip, uh, I was supposed to get back on a plane to India. I was on a tourist visa at this time.

    10. AE

      Mm-hmm.

    11. KD

      Um, and, um, realized, okay, maybe there's some way for me to, uh, continue to be sort of participating in some of these sort of foundation events and just essentially stay.

    12. AE

      Mm-hmm.

    13. KD

      Um, and so then eventually, I guess, uh, came up with the idea of, uh, idea of Lu- Luminai over time and got the... you know, got into YC and, and the rest is history.

    14. AE

      Yeah. Let's talk a little bit about that. Okay. So you came to the US. You were, what, 19-

    15. KD

      That's right

    16. AE

      ... at the time?

    17. KD

      That's exactly right.

  8. 11:00 – 14:07

    Surviving solo at 19: “hacking” hackathons for prize money

    1. AE

      And, um, instead of going back, you decided to stay here, but you're here, you're by yourself. How did you survive during that time?

    2. KD

      So, um, there are these, uh, things that most people know about now, but, but, um, uh, called hackathons, and, um, uh, these hackathons were, like, two-day events where you could essentially, like, uh, build a product.

    3. AE

      Mm-hmm.

    4. KD

      And, um, if you built the best product, then, uh, you could, uh, win up to, like, $10,000 in prize money. And, uh, there were, uh, all these hackers from around, around the Bay Area basic- basically coming to all these hackathons. And I realized, uh, going to my first hackathon, that, um, for me to continue to stay here, [laughs] I need to be able to live and eat food and, and stay somewhere, and, uh, this seemed like a great idea where you could win up to $10,000 a weekend. And, um, so I ended up, uh, along with, uh, a friend of mine, ended up basically devising this, like, plan of how to hack these hackathons. Um, and the thesis was very simple, which was, hey-Who are the actual final decision-makers of who wins the hackathon?

    5. AE

      Mm-hmm.

    6. KD

      Um, what are they wanting out of this hackathon in the first place? And many of them were, like, these corporates who were organizing these hackathons. It was, like, places like Twitter or Google or one of these places, and they had the goal of either recruiting you out of the hackathon, or they had the goal of you testing their APIs. Um, and, uh, what we would do basically is go and figure out what was the core reason they were actually organizing the hackathon in the first place. And then when we got to the hackathon, uh, there were all these organizers who were walking around. Uh, and then basically was the... Maybe the core insight was these organizers basically ended up becoming the final judges who decided who was going to win the hackathon.

    7. AE

      Mm-hmm.

    8. KD

      And so what we would do is basically go to these organizers and show them, uh, what we built in the first two hours of the start of the hackathon. While everybody else was coding away and drinking the Red Bull and doing all these things, first two hours, we would go to them and be like, "What do you think?" And then they would give us some feedback. We'd go back, we'd build something else on top of that. Two hours later, go back to them and be like, "What do you think?" And then we'd do it with everybody.

    9. AE

      Mm-hmm.

    10. KD

      And so by the end of the hackathon, in 48 hours you've gotten, like, maybe, whatever, s- 16 iteration loops with them back to back to back to back. And, um, and you ended up building exactly what the hackathon organizers wanted-

    11. AE

      Mm-hmm

    12. KD

      ... who then eventually became the judges, and the hit rate on that was just ridiculous. And so, uh, ended up living off of hackathon prize money, basically.

    13. AE

      How much money did you win?

    14. KD

      Uh, I think in total our entire group, which was, um, mostly me and my friend i- in a lot of cases, and then added one or two more people, I think total collectively across the maybe five or six hackathons we won, maybe 40, $45,000-

    15. AE

      Wow

    16. KD

      ... or something like that across, uh, the four of us. Um, and so if you split it, it's, like, enough for me to survive for a few months here in the US.

    17. AE

      Yeah. That's incredible.

    18. KD

      Yeah.

  9. 14:07 – 16:41

    Getting into YC with the “wrong” idea—and why YC bet on the founder

    1. AE

      And then talk about, um, the decision to apply to YC. And, and, and you applied with a, a very different idea, not the idea that ultimately became what Luminai is today, right?

    2. KD

      Yeah. So, um, you know, Y- YC, as you can imagine, once you're in Silicon Valley, it's, like, kind of the obvious sort of, um, the most incredible, uh, opportunity, especially for s- someone who, I think at least at that time, felt like this was one of those things where you didn't need SAT scores [laughs] . You didn't need, like, recommendation letters or something.

    3. AE

      Right.

    4. KD

      You were... It felt like the Rubik's Cube thing all over again in some ways, 'cause it was like, it was like the merit of what you are doing right now.

    5. AE

      Yeah.

    6. KD

      And, like, if you're able to show, like, um, if you're able to tell your story and, and, um, connect it back to your vision and what- what's, what's possible, um, then it seemed like there would be an opportunity for me to, you know, um, uh, build a company. And so, a- as all engineers do, uh, the first company idea we applied with, uh, which, uh, we very quickly pivoted away from, was, uh, building documentation for engineers and making it easier for engineers to, uh, create documentation.

    7. AE

      Yeah.

    8. KD

      And so we built basically what today feels like a bunch of companies are doing, and doing exceptionally well, uh, because of LLMs, uh, but, uh, maybe four or five years early.

    9. AE

      Mm-hmm.

    10. KD

      Uh, because it felt like every engineer we talked to was like, "Oh my God, my, my manager's telling me 40% of my time should be spent documenting. I hate documentation. Can you help me solve that?"

    11. AE

      Yeah.

    12. KD

      And so we tried a bunch of stuff and thought this was an excellent idea. We applied with it, and I remember your, your phone call when you told us, uh, that we, we got in that, "Hey, we'd love to take you, but I think your idea's not great." [laughs]

    13. AE

      [laughs]

    14. KD

      "And you should probably reconsider." Um, and, uh, yeah, lo and behold, three weeks later, uh, after our own iteration with the first set of customers we worked with, it became clear it was not a very good idea.

    15. AE

      Yeah.

    16. KD

      Um, and then we went back on the idea maze.

    17. AE

      Yeah. Yeah, it was interesting. I remember that day that we interviewed you. We interviewed four companies that all had the same idea.

    18. KD

      [laughs]

    19. AE

      Um, but, um, you know, I, I think it goes to show that what stood out actually about you was all the things that you mentioned in your background. You know, the setting the world- Rubik's Cube world records, and biking the Silk Road, and coming to the US, surviving on the hackathon prize money. Those were all things that, that in spite of test scores or going to a fancy school or anything like that, all stood out as these are outlier kind of behaviors that are not typical of somebody who's 19 years old.

    20. KD

      Oh, and thank you for taking that bet, because I'm here, uh, because of that bet, so thank you.

    21. AE

      Yeah. Well, it's been amazing to work with you.

    22. KD

      Yeah.

  10. 16:41 – 19:24

    Pivoting methodically: choosing a mission and validating pain

    1. AE

      Talk about, um, the pivoting process and, you know, how you eventually got to the idea that became what Luminai is today.

    2. KD

      Yeah. I think my, my sort of, um, contrarian view on, uh, going through the pivot journey is actually it could be, it can be quite methodical. Um, and what I mean by that is, like, at the end of the day when you're especially building something that is, um, for a, uh, a, a business customer, when you're building a B2B company, it's actually... There's, there's a lot more science than art, is kind of what I've come to learn-

    3. AE

      Mm-hmm

    4. KD

      ... as we've gone through this cycle a few times in many ways. And, um, and so, you know, um, I think there were two threads that kind of maybe we, we thought about at the beginning, which was, um, one, is there a mission or story that truly energizes me to work on a problem for a very long time?

    5. AE

      Mm-hmm.

    6. KD

      Um, and what was interesting was, um, uh, while I was going through this journey, uh, there was a pretty large and serious sort of family health event, and, um, it felt like healthcare felt deeply, deeply meaningful in many ways, um, to me. And then at the same time, as we started to talk to larger and larger enterprise organizations, um, this whole problem of, uh, operational teams mostly doing deeply manual workflows became very evident. Um, and a little random sidebar comment is, um-In, in this whole Rubik's Cube world, if you think about how to actually get faster, it's like how do you actually cut the number of steps it takes to-

    7. AE

      Mm-hmm

    8. KD

      ... solve a Rubik's Cube?

    9. AE

      Mm-hmm.

    10. KD

      Is like one of the methods that you get faster. And I was actually an incredibly inefficient solver growing up. Like, meaning, um, the reason why I got very fast was because my, like, dexterity from like a speed of, um, the actual Rubik's Cube, um, turns was faster than most people-

    11. AE

      Mm-hmm

    12. KD

      ... uh, in the world. And so when people would l- see my solves, the first thing they'd say, they'd say is like, "How do we cut the number of steps-

    13. AE

      [laughs]

    14. KD

      ... to get you to a solved state?"

    15. AE

      Mm-hmm.

    16. KD

      Um, and, uh, and so when you think about for some reason this, uh, this problem of, like, how do you cut the number of steps from someone's workflow felt, funnily enough, attractive. Um, and, uh, and so when you start sort of going down both of those threads, of one, hey, is this a mission or a thing that I actually care about, and then you actually go deep and talk to a bunch of different users and customers to understand where's the biggest sort of pain, and then talk to enough of them that, like, you start thinking about what is the generalized problem they're sort of talking about, um, you start, you start to actually pull threads that eventually lead to the creation of a, of a real company.

  11. 19:24 – 23:39

    Enterprise sales at 20: selling a champion, not the institution

    1. AE

      Mm-hmm. And you mentioned, um, selling to enterprises. I think one of the things that stands out to me is you were closing big enterprise customers at, like, 20 years old. Talk about that. Like, how were you able to do that? I, I, I know a lot of 20-year-olds, I think, are self-conscious about that or thinking, you know, "H- how can I talk to these big corporations?" And, um, what were some of the tips or tricks that you had in the early days that worked really well for you?

    2. KD

      Yeah. I think a, a couple of things. So one is, um, what's, uh, really, really sort of maybe, um, not clear from the outside in is, um, that in the end of the day, you're not actually selling the enterprise, you're actually selling an individual.

    3. AE

      Mm-hmm.

    4. KD

      I think that was the first very clear insight, or the cl- the thing that became very apparent to me is-

    5. AE

      Mm-hmm

    6. KD

      ... you're not actually selling a customer like Cleveland Clinic, you're selling a champion-

    7. AE

      Mm-hmm

    8. KD

      ... within that one institution who has to say, "Yes, I want to work with this product or team."

    9. AE

      Mm-hmm.

    10. KD

      And that became maybe the core part of the strategy, which was like how do we get in front of these people who are forward-leaning, uh, meaning people who would be willing to take these bets and actually maybe think differently than the rest of the organization. Some very small signs we saw was, like, someone just, like, growing rapidly through their, like, career-

    11. AE

      Mm-hmm

    12. KD

      ... within that institution.

    13. AE

      Hmm.

    14. KD

      There would be people sometimes in a hospital who've been there and they're in their 30s, but they're like a C-suite executive-

    15. AE

      Mm-hmm

    16. KD

      ... at the hospital. That's incredibly hard-

    17. AE

      Mm-hmm

    18. KD

      ... to pull off in that type of bureaucratic, like, 100-plus year institution.

    19. AE

      Mm-hmm.

    20. KD

      And so what we tried to do is basically look at those types of folks and then figure out who are the people that they know that we might know second or third degree-

    21. AE

      Mm-hmm

    22. KD

      ... sort of a way. Um, activated basically anybody and everyone who would be willing to sort of help. And, um, we would basically call up each of these people and say, um, "Hey, I'm, I'm gonna send you, like, our latest, like, uh, value proposition. Would you just be willing to say a ni- a few nice words about Illuminai and me," and ask them if they'd be willing to meet with me. And in a lot of cases, because people just kind of self-select out of reaching out to these people, um, w- we just got in front of so many, like, really, really top executives at a lot of these major institutions. And then once you get there, you actually, funnily enough, absolutely, at the end of the day, if your product doesn't have value, you're not gonna... there, there's not gonna be a sale that happens.

    23. AE

      Yeah.

    24. KD

      But one of the m- number one qualifying criterias, which was a real surprise to me, was when you lead with your personal story and when you lead with your, like, um, your history in life and you lead with your mission and the mission that, like, the company is working for- toward, uh, even if you haven't remotely accomplished it, uh, they buy into that narrative.

    25. AE

      Mm-hmm.

    26. KD

      If they believe it and they want to take a bet on you, that's how these, like, partnerships get done.

    27. AE

      Mm-hmm.

    28. KD

      So I would say, like, if you think about some of the major customers we've gotten the chance to work with, most of them, um, were because we were able to tell a, like, personal story that really got their attention, um, whether it was mine or my team's or anybody else who, who was at Illuminai that really got them interested and connected to want to do something special. I would say it's, in the end of the day, like a deeply personal, uh, journey [laughs] uh, uh, called enterprise sales. [laughs]

    29. AE

      Yeah.

    30. KD

      It's not, not what maybe outside in I, I learned.

  12. 23:39 – 26:51

    Warm-intro pipeline engineering: scraping networks and climbing the org chart

    1. AE

      Mm-hmm. For the first few years of the company, you didn't send cold emails to get these customers. It was all through introductions and network and friend of a friend and that type of thing that you were working, right? I mean, th- that's probably counter to how a lot of people think about how they should start with sales, which is, okay, I look up all these people on LinkedIn, and then I send them a bunch of, you know, spammy messages and see who responds and go with the one person that responds. But you took a very different approach. Like, talk a little bit about that, how long you did that for.

    2. KD

      Yeah. I mean, um, uh, you know, it's kinda funny because I do think, like, probably, like-Some major portion of our customers still come through that, that way. The reality is, like, really major institutions, uh, don't, don't really talk to you unless it's someone who they trust is, like, making the referral.

    3. AE

      Mm-hmm.

    4. KD

      You know, back in the day, obviously it's changed now, even if you look at, like, venture capital, they used to always talk about, like, "I need a warm intro. I need a warm intro."

    5. AE

      Right.

    6. KD

      Large enterprise customers are also not that different. Um, and so, uh, and I'll tell you, I'm probably living proof of this in one way, which is it's not like I had a network.

    7. AE

      Mm-hmm.

    8. KD

      You know? It's not like I showed up with, like, a Rolodex of all these people. So I think, like, what was, what was really, um, what was really interesting was, like, um, we would just lean on saying, "Okay, who is most likely to be connected to this person? And who is that person most likely connected to that we might potentially know?" And then basically just sending out, um, you know, uh ... And it, and it took time and effort, right? Where you, like, actually said, uh, to maybe even your small investor crew at that time, "Here's the, like, 13 people I'm trying to get, get in front of."

    9. AE

      Mm-hmm.

    10. KD

      We wrote a little script where-

    11. AE

      Mm-hmm

    12. KD

      ... we would, um, and you've received a bunch of these [laughs] notes.

    13. AE

      Mm-hmm. [laughs]

    14. KD

      But what we d- wrote a little script where we would go through everyone's LinkedIn and scrape every single contact of theirs-

    15. AE

      Right

    16. KD

      ... and then just match it against our, like, target customer profile.

    17. AE

      Mm-hmm.

    18. KD

      And out would come maybe two or three names, and then those two or three names, what we'd do is be like, um, uh, "Hey, Aaron," like, uh, "here are the two or three names. Do you actually know them?"

    19. AE

      Mm-hmm.

    20. KD

      And then they might be like, "Okay, I know one of the three."

    21. AE

      Yeah.

    22. KD

      Um, and then we'd send a, like, short blurb. It actually didn't matter for us that much if they were executives or anything like that. They just, if it was a target customer we really wanted to get in front of, I was like, if I can just find someone at the company, I'm sure that it, we can pitch their, pitch it to them and get them excited, that then they might be willing to get us maybe two layers above, and then slowly make our way to the actual person who's the champion.

    23. AE

      Right.

    24. KD

      But if you do that iteratively over many, many, many, many, uh, weeks, um, and months, um, you'll have a very, very thick, uh, pipeline-

    25. AE

      Mm-hmm

    26. KD

      ... of people who are willing to have deep conversations with you, to partner with you, and, and, uh, that's still a major, uh, part of the current, um, uh, cycle of, uh, enterprise partnerships we do. Obviously the network has changed. It's no longer investors. It's, like, customers making the referral, or, you know, uh, advisors who are, like, big, um, you know, uh, uh, big-time, really serious people in the industry that have now starting to champion Luminai.

    27. AE

      Mm-hmm.

    28. KD

      Yeah, um, and so that's changed. But by the end of the day, like, this is a, like, deep relational game, I think, at the end of the day. And, and the world is much smaller, uh, especially if you focus your customer base, uh, than, than I expected it to be.

  13. 26:51 – 29:50

    Reinventing the company: from horizontal automation to healthcare focus and clear wedges

    1. AE

      Yeah. And I mean, you used that strategy to grow to many millions in revenue, and then you had to reinvent your, yourself and, and the company and the product a little bit. Talk, talk a little bit more about that.

    2. KD

      Yeah. So, um, Luminai, uh, was a much more generalized and horizontal automation platform. And, um, I think what sort of started to become clear, even after you start to get, like, you know, dozens of customers and many millions in revenue, um, that you actually have to essentially, um, become a really important partner to, uh, a very clear set of customers. And you can't do everything for everybody. And, uh, that realization hits when the next customer you talk to, uh, especially as you go larger and larger enterprise, um, uh, the moment you start to say, "I can do anything," or, "I work with someone else who doesn't look like you at all," the credibility you have reduces pretty dramatically. And what was interesting is when you looked at the data of our existing customer base, um, it was, like, north of 80% of our customers were in healthcare.

    3. AE

      Mm-hmm.

    4. KD

      Um, and, uh, we made the big sort of decision to say, "Hey, uh, we can't try to be everything for everybody. Let's focus in and say let's go all in on healthcare." And so that was the first maybe decision. The second was being a horizontal you-can-do-anything-for-everybody in healthcare is also a, like, you know, uh, in some ways a, uh, a problematic setup to grow into the type of company we wanted to really go out and be. Um, and so we started to say, "Okay, what are the actual clear, thick use cases that we can actually drive a ton of value in?" And, you know, it's like these institutions are large, and they don't have time to think about what they can do creatively with your platform. You have to almost tell the story of what's possible very, very clearly.

    5. AE

      Mm-hmm.

    6. KD

      And you have to understand their problem so deeply that you're able to talk to their problem like you've known them for 20 years. Um, and so, uh, what we started to do is continue to be sort of extendable to do many other things, but then have a very clear set of initial starting use cases and wedges that makes them realize value incredibly quickly.

    7. AE

      Mm-hmm.

    8. KD

      And so that's the transformation we've, we've gone through as the company's grown and, and, um, gotten to a place now where, uh, I'm sure there will be continued transformation as we make, uh, make, uh, the next set of moves as we become larger and larger and larger. But, um, and we scale our impact from a value perspective with each of these customers. But, um, the reali- the realization I've had is just, like, with every stage you have to reinvent yourself.

    9. AE

      Mm-hmm.

    10. KD

      Um, uh, because the truth is, like, the next YC company's gonna come in and pitch the most, uh, the, the next, uh, dreamy vision-

    11. AE

      Mm

    12. KD

      ... that, that we were in five, four years ago.

    13. AE

      Yeah.

    14. KD

      And, um, you have to become that, uh, while scaling your, your, your company. And so that's kind of what, uh, you know, at least we've realized, uh, as we've grown. And, and, uh, honestly it is a deeply energizing experience in many ways as you transform and reinvent and, and continue to compound on all the learnings you've, you've had. And that's definitely happened-

    15. AE

      Yeah

    16. KD

      ... as we've, as we've, uh, built the company.

  14. 29:50 – 32:51

    Founder advice: scale your ambition—and prove it with intensity

    1. AE

      Yeah. I'm curious, if you had to give one piece of advice to yourself, maybe something that you wish you heard when you were 19, came here to America, were starting to think about starting a company, what's the thing that you wish you heard back at that point or that, that you would've liked to give yourself?

    2. KD

      The level of aggression and ambition you can have is probably 100 to 1,000 times larger than you think you should have.

    3. AE

      Mm-hmm.

    4. KD

      And it has nothing to do with, like, your experience or, uh, where you come from. There are, like, no rules [laughs] ... is maybe the, I mean, there's obviously integrity rules. I think you have to, uh, but outside of that, I don't think there's any real rules, um, with, with the types of problems you can go out and, and solve, and that you have access to, and, and transform. And, and, uh, I think especially now in the era we're in, um, I think, uh, it's becoming very evident that, uh, the playing field has never been more level. Um, and, and so yeah, if you want to become the next, I don't know, Steve Jobs [laughs] or Bill Gates or, or Elon Musk or whoever you look up to, I think it's maybe much more possible than, um, maybe I would've dreamed of. And I think what the other interesting aspect of it is, uh, I think I have the same level of intensity, maybe significantly more intensity now, and I'm sure my like 20-year-old self would be very happy with where we are right now, and, you know, we've raised about $60 million and all of these things. But I think now the, the level of sort of pursuit for the mission feels multi-decade and very lifetime versus, um, something that, um, was maybe much smaller when I, when I started. And I wish if I was m- maybe more ambitious, maybe would've gotten to this position significantly faster.

    5. AE

      [laughs]

    6. KD

      Maybe. Um, but, um, yeah, I do think that like it's been, it's been helpful to, uh, dream like anything is possible.

    7. AE

      Mm-hmm. Yeah. You're living proof of that.

    8. KD

      Oh, thank you.

    9. AE

      And where you came from, and teaching yourself Rubik's Cube, that, you know, just keep stacking these things one on top of, of another until you... Now you're at the point where you've raised $60 million, and you've got this company serving all these huge healthcare clinics. And, you know, we'll do this again in five years and see all the incredible things that you go on to do from here.

    10. KD

      Awesome.

    11. AE

      So.

    12. KD

      Thank you, Aaron.

    13. AE

      Yeah, appreciate you joining, um, to tell your story. Uh, it's always inspirational to hear. Amazing. And I can't let you leave without having you solve the Rubik's Cube.

    14. KD

      That sounds great.

    15. AE

      Okay.

    16. KD

      I'd love to.

    17. AE

      All right. I'm gonna s- I'm gonna mess it up for you. Ready? Go. [Rubik's Cube crunching] What? 10.58 seconds.

    18. KD

      [laughs] Whoa. Amazing.

    19. AE

      That's incredible. Wow.

    20. KD

      All right.

    21. AE

      Keshav-

    22. KD

      Thanks

    23. AE

      ... thanks so much for joining us. This was amazing.

    24. KD

      Well, thank you for having me.

    25. AE

      Yeah, appreciate it. [upbeat music]

Episode duration: 32:51

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode 00YZWlvAMsQ

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

iOSAndroidClaudeChrome