Full Transcript

Dhruv Sharma: Hey there listeners, this is stream number 118 and before we introduce our guest for today, we're just out with our fourth roundtable in space where we have three space entrepreneurs and one front-end tech investor. Please take a look at it if you haven't already. Today's guest is Srijesh Kumar. Srijesh is the CPTO, the Chief Product and Technology Officer at Practo. You'll remember the OG health tech company that all of us have used at some point in our lives. Srijesh, welcome to The Offline Network.

Srijesh Kumar (Practo): Thank you Dhruv, excited to be here.

Utsav Somani: So Srijesh, you've of course had a very storied career even before you came up to Practo. Talk to us a little bit about your journey up to this point.

Srijesh Kumar (Practo): Sure, sure, sure. Yeah, I've been working since 2002 and like everyone else in that era, I pursued my typical com science bachelor's and then got into the tech sector. And since then, I've been basically building a lot of products. I've worked for Adobe, Microsoft, then I did e-commerce with Expedia. Then I had this big thing to go work for a SaaS product. So then I moved on to a startup, I headed the engineering there. And then I moved on to doing SaaS at a much higher scale with Salesforce. And then, I wanted to go back and do something more impactful, meaningful, etc. So that's how Practo happened. So I'm excited to be here.

Utsav Somani: Srijesh, I'm curious, how has technology and product leadership changed in the face of this AI super cycle from your perspective? I mean, so obviously, there is a lens that you have to put when it comes to everything that you do. One of the things is, hey, how can AI make whatever you're doing easier, simpler, and better? Not just for our customers, obviously, you have to have that customer lens, but also for the internal team members, developers, product managers, our sales people, our operations people, etc.

Srijesh Kumar (Practo): So the ecosystem is... And there is a potential to do a lot of things quickly as well. And so it's still like, you know, peeling the onion. And we're fairly young in that journey at this moment. But there's a long way for us to go at this moment in terms of what we can do with AI.

Dhruv Sharma: You mentioned that you're four months into this role. And you've inherited a product, which is built over 18 years, Practo is 18 years old, I believe, and live in over 20 plus countries. You have a consumer facing business, you have a caregiver facing software as well. So when you took a look from the outside to the inside, what was the first thing that you said that this needs to be AI enabled or this can be touched by AI and improved by AI?

Srijesh Kumar (Practo): I mean, when I looked at it from outside, I was thinking more about patient experience, right? And the experience that I as a Practo user had when I would go to a clinic or a hospital, right? And I wanted to simplify that. And I always knew that when you visit a clinic, so many things that a doctor or their assistant or the panel has to do, which is not truly talking to you or understanding the problem that you have, which is basically ancillary to their core piece of work. For me, one thing that I really wanted to do was like, how can I make these doctors essentially spend more time with me than writing on a piece of paper or, you know, doing all this ancillary stuff and like, you know, how we can get to that diagnosis faster. I think that was one thing that I really wanted to change. And luckily, I've ended up at a space where we can actually, I can make a difference, to be honest.

Utsav Somani: Yeah. And Sushant mentioned this term a lot, care navigation. I think that's a very loaded term. So where does it, I mean, unpack that for us? And also, where does Practo's role come in? Like, do you basically, what's the hard part of it? Is it the search? Is it the ranking? Is it building the trust when you're navigating something as delicate as healthcare?

Srijesh Kumar (Practo): Yeah, it's probably, it's all of it. So when you look at it, right, let's say that there are two facets to this entire problem statement, right? Like there is a care seeker. The care seeker could be someone who has an illness or who needs some kind of medical help. And then there is a care provider, which could be a doctor, a clinic, or a hospital or something, right? So the journey begins when somebody has to, like, you know, has to go find a doctor for a particular problem statement. From helping them discover which doctor to go to, right, to find that care that they need, to essentially handholding them all the way till their end outcome is achieved, right? That entire journey is what we call care navigation, which is navigating two different milestones when you are trying to get that care.

Dhruv Sharma: I think you're one of the only companies to display recovery rate as a metric, if I'm not mistaken.

Srijesh Kumar (Practo): Right. I mean, yeah, so we are also like, you know, now we also are looking at, you know, see, our eventual vision is basically that we want everybody to live healthier, longer, and that's the goal. And which is why all of us are doing what we are doing here. Recovery rate, or as we call it, like a health outcome, right? It's very important, right? Our job doesn't end when a customer books a particular appointment with a particular doctor, right? It only begins there, right? So which is why we've been trying to understand more and more about like, you know, what happened when a person actually visited a doctor? What was the outcome of that discussion? How did that go? What feedback do they have about the doctor and the whole experience? And then from there, if they need like, if the doctor prescribed some medicines or some lab tests, or let's say some kind of IP procedure or surgery, like we want to understand that too, and then ensure that, you know, let's hypothetically say that, hey, they've been prescribed a surgery, we want to handhold them and then, you know, take them to a place where they are able to discover the right place to do the surgery at the right price, at the right time. And then also, we will go ahead and help them schedule that surgery at the center. And then post surgery, we also try to understand what happened, how are they feeling, etc, etc. That it's so simply put, it doesn't end by just visiting the clinic, it only begins there, to be very honest.

Dhruv Sharma: Fantastic. Sorry, go.

Srijesh Kumar (Practo): No, no, also, please go ahead.

Utsav Somani: So I mean, the public, I mean, infrastructure is also working, they're building on these public goods, they've done for Aadhaar, they're also doing for healthcare as well, like the ABHA ID or the Smart Doctor by AIIMS, which is like a decision support system for taking these decisions, which is I believe, is rule based, not AI based yet. So how do you build as a private player in healthcare on top of this? Isn't that something that Practo is exploring?

Srijesh Kumar (Practo): Yeah, with ABDM and the whole Ayushman Bharat thing, like, I think, we are connected, we have done some work in that direction. But even beyond that, right, I think one of the things that we need to understand is nobody has done this, what Practo is doing in the last 18 years, there is no one literally no one who has done this in India, the level of depth that we have, honestly, what that allows us to do is like, you know, we are thinking about collaborating with Frontier Labs to build out a clinical decision system. So we are currently having some ongoing conversations because there is no one better to actually do other than Practo based on whatever longitudinal data that we have about patient care.

Dhruv Sharma: Yeah, I mean, and that actually makes a lot of sense, Srijesh, and Practo is uniquely positioned to execute on that. I mean, there's this one company I track in the US called Open Evidence, where you would know this company where, I mean, charity for medical purposes best used by patients, not so much by doctors. So they really are, I mean, the training data for Open Evidence is, you know, journals and scientific literature that's typically behind a paywall. So that's that's another example of a great company, just like Practo. Questions for you, Srijesh, one is, I've actually gotten very interested in the fact that you're about four months into your tenure. So the question is, when, you know, when a senior leader like you comes into a large, you might say, a fairly mature organization, what are those first three or four months like? How are you assimilating? How much are you listening versus speaking? When is the right time for you to start sharing, you know, the experience you've accumulated over many, many years? How have you been navigating that?

Srijesh Kumar (Practo): Yeah, I mean, so one of the things, like, you know, at the seniority, one of the things that any leader should have as a skill is, like you rightly put out, listening more and speaking less to begin with, right? Like, you know, so I think a fair part of last four months for me have been spent understanding, you know, the state of union, where Practo is, like, you know, what the, where we are headed, and so on and so forth. And it, you have to basically unpack a lot of stuff. And then probably come up with, like, a playbook on, you know, what it would take to actually, you know, execute on the vision that we have. So I would say, like, you know, I need to continue to be in that listening mode. It's just the mix that changes. Initially, if it was like, you know, hey, I was listening 95% of the time. Now I'm also, maybe that mix has come down to 50%, right? Like, and so, and that mix might just change. It's a slider. But listening doesn't stop, right? Like, for as long as you are, you're doing something interesting. So whether it's not, and it's not just about the company, it's about also listening to your customers, doctors, your partners. Yeah.

Utsav Somani: And Practo, for the first time, I think is unified product engineering and data under you. So what part of customer journey is now enhanced because of that decision? Or what do you think was not working at Practo before in terms of internally, like the breakages between different siloed units within the company that are now enabled because they've clubbed this together under you, your leadership?

Srijesh Kumar (Practo): Yeah, I share two perspectives here. One is the larger perspective, right? Like, you know, so it's not, what we have done is not unique to Practo, to be very honest, right? Like this whole thing about having one person accountable versus having, you know, a function specific specialization, it's a pendulum. Basically, organizations keep swinging in from one to another. And one of the reasons is like, it's all about what you need at that point in time, right? So for example, what does unifying all these organization help Practo does? It helps Practo put one person accountable for the entire product execution and that the buck stops at me, right? Like, so there are no excuses for lack of collaboration, for lack of execution speed, for lack of decision making, end of the day, like, you know, yeah, I mean, there is that velocity that has been built because of that accountability, right? That exists with me. That being said, like, you know, there may be organizations out there, which are probably wanting to go deeper into a particular discipline like user experience, right? So, so they may take a very different perspective that they want a separate vertical reporting all the way up to CEO, etc, etc. So, so like, so yeah, I mean, the gist of it is, it is what the organization requires. And right now, Practo needs that velocity and that accountability. And that's why it's not about what we did earlier was wrong. It's more about what we need to accelerate in the future.

Utsav Somani: And you've reported profitability as well for the first time.

Srijesh Kumar (Practo): We've been profitable for I think, nine out of seven quarters last nine out of seven quarters or something. So yeah, so I think, yeah. Thank you. Thank you.

Dhruv Sharma: Thank you. One of the things we haven't had a chance to touch just yet, Srijesh, is like the recent international expansion at Practo. Talk to us a little bit about that journey and the learning so far.

Srijesh Kumar (Practo): Yeah, so I mean, different markets, different requirements. So I mean, currently, you know, we have footprint in UAE. We also have a footprint in the US. We're very, very young in the US, to be very honest, we've built out the product. But we are not accelerating on it at this moment because of certain reasons. But we will be expanding there in days to come, if not months. UAE has been very interesting. We already had a significant presence in UAE because of our B2B business, our hospital management software, etc. And it's fairly popular there. So for us, you know, expanding in UAE was kind of the logical next thing to do from our B2C product perspective. There are some local nuances. I think the bigger thing is we have built out a product which is, which is kind of global in nature. There are certain things that you have to tweak to be locally relevant, right? Like, for example, US, if you take that, it's very compliance heavy, right? Like in terms of like, I think, it's very insurance heavy as well, right?

Dhruv Sharma: Healthcare systems are entirely different in the two markets.

Srijesh Kumar (Practo): Correct, correct, correct. So you just have to worry more about those nuances that make us more locally relevant. And yeah, and that's, that's where we have to invest if we are to go global.

Utsav Somani: And I think there was a statistic that I read online where 60% of the code now at Practo is written by AI. So I mean, what has changed in the team? Like, are you hiring less or AI specific engineers? What do you look for as you build out your team now at Practo?

Srijesh Kumar (Practo): We're doing both basically. So I think we already had a lot of smart engineers here, right? Like, and so I mean, honestly speaking within the company, I mean, the Uber CEO famously said that they ran out of their AI budget very quickly. We do, we do, we do, we do have spend gaps. So yeah, I mean, I manage that budget. So I cannot cannot go wrong there. So it's already very particular about spend management there. The, the thing is there are two problem statements, right? So one is, hey, how do we pivot the existing team members? They're very smart people. They understand this ecosystem. So we put an effort to essentially pivot them, which is more about like, you know, building the right kind of harnesses, the support system where they can start using more and more of AI. I'm surprised with that 60% stats, because my understanding is probably 80, 90% of the code today is written by AI. Where we need people to spend time is probably things that are, that are to be done after writing that code at this moment. And the second thing is around AI engineers. We're also doing, hiring a few people and, you know, blending our team with the AI skills that existed out there in the market. Primarily because our eventual goal is to not build some kind of siloed AI team. Our eventual goal is that our entire engineering product organization, and even beyond that becomes like an AI first organization. And the idea is that these people can probably build out the platforms that will enable us to do so. For example, a sales executive sitting somewhere should have a platform eventually where they can say that, hey, I'm going to build up an agent, essentially automate my job and, and, you know, and essentially make me move, help me move faster, basically. So that's, that's the goal. That's the ultimate vision.

Dhruv Sharma: I think two quick follow ups there, Srijesh. One is like, there's this, there's this debate that's been going on for a while on agents, which is what does one begin with like hyper narrow, hyper specialized agents? Or do you start, you know, very early on thinking about a more general purpose agent? That's number one. The second is, in light of everything that you said, are you actually measuring AI fluency across the org? Are you rating people for it from a, you know, from a talent management standpoint?

Srijesh Kumar (Practo): Yeah, absolutely. So the second question is probably easier to answer. So I'll just take that first. So we are measuring AI fluency. And it's a, it's a metric that we basically want to continue to improve over next few months. So that is that. The other thing coming back to,

Dhruv Sharma: can we go slightly deeper on that Srijesh, so that if there's someone else considering bringing that into their organization, also, they have a method on how to do that.

Srijesh Kumar (Practo): Yeah, yeah. I mean, see, I mean, we are, we are very deep in terms of measuring our AI outputs. We are looking at the PRs that are written by the lines of code that are being around there. And then, you know, also asking this question across our business processes as to like, you know, hey, like, you know, if you are doing this today, and it takes you X days, you know, what can we do to bring that time down? And, and therefore, that is involved in getting this done, etc, etc. So all those things being measured at this moment, in many ways, and then we have come up with a composite blended metric on AI savviness, which we measure. And the goal is to essentially improve that metric.

Utsav Somani: Excuse me, sorry. Yeah, so I think one, another thing that I want to probably like, pick your brains on is that put on your product hat, like I mean, many founders, potential founders for the healthcare ecosystem are listening in, I think maybe we can expand and the next two, three minutes just learning about what the digital public rails look like for the healthcare ecosystem, because a lot is happening. And a lot of people are not aware, because probably the government doesn't do a good job of marketing them. But a lot is getting built out data sets, I mean, support decision systems and many other things. So maybe we'll take that.

Srijesh Kumar (Practo): Sure, see at a very high level. I mean, the big tech giants at this moment are focused on going deep on health tech, and amongst many things. And I think clinical diagnosis is which is something that they are going after. And there's a lot of model training that has been happening in that space specifically, so that you know, they can understand the health outcomes more. And, you know, and probably, you know, test out their models, run some events, etc, on top of it, etc. So there's a lot of work research work around that, that's happening in the frontier lab space at this moment. Now, specifically speaking within Practo, right, our goal is not to replace a doctor, which may be an intent for some other organizations, somewhere, our goal is to ensure that a patient and a doctor spend more time with each other, and then healthcare is more available, right. And there are faster and better health outcomes out of that interaction. So we are, we're not building systems which will, which will replace a doctor, we're not building systems, which will replace a clinic, etc. We're not focused on that. Our goal is that, hey, we believe that doctors are important. And it's important for patients to connect to the right doctor. So we want to build systems that will ensure that that conversation is more fruitful. And that eventually leads to, you know, better outcomes for the patient. One of the examples is we've built out like a OPD concierge product. What that does is we call that agent Ali. So what Ali does is essentially Ali, you know, the moment you are, you book an appointment, Ali will, A, Ali will also help you book an appointment. If you call up Practo, then, you know, Ali would send you reminders, nudge you for the, for the clinic walk-in. He'll interact with you over WhatsApp, and he'll talk to you on the phone. And once you're, once you're done interacting, he'll collect the feedback, understand how your experience was. And based on that, that decides the next steps that are there. And, you know, if, let's say, I spoke about this example earlier, if a surgery is required, Ali will probably help you schedule that surgery as well. So, and eventually once, whatever your eventual outcome is, Ali will try to measure how that experience was and whether that outcome is achieved or not. So that entire concierge is like an autonomous agent at this moment that we have built out. And for us, for Practo, that is the space that we want to be. We're not building like, you know, complex systems that will do medical diagnosis.

Utsav Somani: But I mean, the government initiatives like ABHA and National Health ID and like other things that maybe we can do a quick listing of them. I mean, there are, there are so many of them, I mean, that, that are there.

Srijesh Kumar (Practo): So obviously, we have been working on the, on the ABHA thing for sure. Like, you know, and that integration exists within Practo at this moment. We've also done a lot of, you know, I think government has also mandated a certain level of accessibility, compliance, et cetera, that has to exist in the, in the health tech space. So we are also investing in that direction heavily at this moment. So yeah, those initiatives are there. I am not, I'm not super qualified to probably answer those or speak in depth about those, but, but yeah, I mean, whatever.

Dhruv Sharma: Maybe just one way to, I'm sorry, maybe one way to square that off would be just like in India, we've had the government do a lot of things that traditionally the private sector has done in, in, in other countries or in like in Western countries, like for instance, the payment rails over here, public utility is, is the vision with Ayush, I think is, also Ayush is the name of this, of, of.

Utsav Somani: There are a few others like ABHA ID and a bunch of others as well, including the system.

Dhruv Sharma: Do you both think that the vision is to kind of create like EMRs, like electronic medical records for all of India and, and then do a, build a lot of applications on top of that? Is that the grand vision that they have?

Srijesh Kumar (Practo): That is correct. I mean, the idea is to have one system for it, for patient records of EMRs, like you rightly said, right? Like, so that's the, that's the goal. And then you have that data.

Utsav Somani: Ayush I think was the Ayurvedic one, just read up, ABHA is the national health ID. ABHA is the one, ABHA is the national health ID. You are right. Yeah, yeah, yeah.

Srijesh Kumar (Practo): So, so yeah, I mean, the idea is to essentially have a Aadhaar equivalent of EMRs basically. And so, and then there are different applications that, that particular repository can bring in. I mean, let's talk about insurance as an example, right? Like, you know, I think the insurance approvals can be faster, better using all that data. If you, if somebody has access to that, that data early enough. But I mean, at a very high level, this health tech is a very fragmented system at this moment, right? Like it's a very difficult problem to solve. So I think every player in the health tech space has to contribute and make that system successful so that the network effects of that come back to us as well, right? Like, and I think everybody collectively benefits out of it, having that one unified system.

Utsav Somani: Thanks for sharing, Srijesh. Yes.

Dhruv Sharma: Yeah, I think you were talking about systems and software. I think we've discussed that aspect, but I believe Practo runs co-branded clinics as well, 40 plus of them.

Srijesh Kumar (Practo): So that's right.

Dhruv Sharma: So tell us a little bit about the physical stuff.

Srijesh Kumar (Practo): Yeah, I think what we do is we, there are clinics out there, which Practo helps manage basically with our expertise. And it's one part of our business where we essentially, I mean, financially put, we run the books, we own the books, and we also own the software that powers those clinics, essentially. We have all the things, except what we didn't have is like a physical location. So this kind of a business model helps us, you know, bridge that gap and continue to offer, you know, healthcare. You know, how do I put it like, you know, there are certain specialties where, where we believe that we do really well. Yeah, yeah, there is a, and there is a gap and we can probably make a difference. And that's, that's what we are going after. There's a lot of room for us to scale in that direction. At this moment, it's, it's a, it's a young business.

Utsav Somani: I think the whole integration in healthcare has been fascinating to watch. Like, I mean, Policy Bazaar is of course a big one, like even which was a healthcare service provider, insurance provider, I think, also started setting up their own clinics. So I think a lot of people are going backwards just to give better healthcare. And there's one hospital, I think, which came up in Bangalore called Super Health, which apparently is delivering a phenomenal customer experience as well. So it's amazing to see that people are doing many different directions in the healthcare space to vertically integrate and also give a better patient experience. Yeah.

Srijesh Kumar (Practo): The way to round this up is it's a very complex space, to be very honest. And it's very fragmented too, in terms of, so all these efforts to bring it together, you know, go deeper into specialties and, you know, et cetera, et cetera, are all moves in the right direction, but there is still a long way to go.

Dhruv Sharma: By the way, I think Super Health is, I think the co-founder is X-Practo. Oh, wow. Nice. Well, that's another interesting thing to think about as I'm assuming Practo's also built its own mafia out of in the healthcare domain over the last many, many years.

Srijesh Kumar (Practo): Yeah. Yeah. I think we had a foundation state where we had some of these guys come in, I think two of them, but a lot of people who have left Practo have gone and found their own companies in the health experts. So it's just, it's a very, very unique thing to see, to be very honest, I think.

Dhruv Sharma: Yeah. Srijesh, I think we're going to close with an audience question. Alexander has asked us on a chat, product related, if someone is early in his career, how to work on improving product thinking in the era of AI, any books slash resources, if you could share.

Srijesh Kumar (Practo): I mean, resources, I mean, I wouldn't recommend books. I don't think the best thing in my opinion is to like get one of these, you know, Claude Code or codex licenses and just start building an idea that you have in mind and you learn, learn as you build it. And yeah, I think that's probably the best way to learn, according to me, for me to start.

Dhruv Sharma: The tool itself teaches you actually how to use it.

Srijesh Kumar (Practo): Absolutely, absolutely. Yeah, absolutely. So, I mean, I think that the world has changed. We probably don't need a lot of tutorials and trainings to get started. It's these tools are fairly simple and easy to use. And you should start by building and just keep building. That's the best way to learn.

Dhruv Sharma: Are you on the side of the AI optimist or AI doer? I think there was a lot of open letters and even an op-ed by Mark Zuckerberg, which was posted like, you don't have to take a stance publicly, but website.

Srijesh Kumar (Practo): No, no, I'm a, I'm a AI evangelist. That's how I would put it. And it's, it's been a huge blessing, honestly. I think we can, there is, we can move fast. We can really make a difference if used the right way.

Dhruv Sharma: Amazing. Srijesh, thank you so much for coming on our show. Wishing you all the best in this new role.

Srijesh Kumar (Practo): Thank you Utsav, thank you Dhruv. It was wonderful chatting with you guys.

Utsav Somani: Thank you so much for coming.

Dhruv Sharma: All right, listeners. We don't have a second guest today. He couldn't make it. So I think that's it from us. We'll see you on Friday for stream number 119. Have a wonderful rest of the week. Thank you for tuning in. Bye-bye. See you on Friday.