Supreet Deshpande, Cintio Labs: Agentic and voice AI for pharma commercial teams

Welcome to another YC Wisdom podcast. We have Supre Desh Pande with us today. Supriit is co-founder and CEO at uh Cintio Labs, a YC company I believe.

Yeah, one of the latest batch, right? Uh 24 I would say [snorts] >> 25. Okay.

So even even better. Uh JP Morgan, XJP Morgan, X McKenzie, right? So and also it's not your first company.

I think first one was like uh Winkle >> and that was like 10 years ago so a long time ago but >> yeah dude nice to have you on tell us a bit more about yourself and what you've been up to >> thanks Charles no really really excited to be here u yeah so I've you [snorts] know grew up in India um always wanted to build something so that's sort of my first venture which was back in school uh which which was a fun exercise but somehow ended up in in the consulting world. So did that for a couple years in India then moved to US uh sometime in 2019. Um back then there was no AI so it was all very machine learning or data science.

So so did my masters there and joined Mckenzie right after. So predominantly worked with a lot of life sciences and pharma companies. So that's been my world ever for for forever.

Um while I was doing that there you know the whole GPT thing happened in 22 and then everyone was like hey what do we do with this new piece of tech right and and that's when given I was always in in the AI side of the house McKenzie put together a small team called geni for life sciences so I was involved there so got to kind of explore some of these early bursts of startups and then had a unique opportunity at JP Morgan as well where I was directly leading the Genai strategy which was essentially any startup that wanted to sell into JP Morgan had to kind of go through our team and so had almost like this front row seat into you know call it scale AI harvei some of these early agenti companies being built so that's when it became very clear that hey this is this is a great time to build something so yeah so that's in in some sense my background uh applied to YC got in I was on the East Coast uh the last seven eight years of my life moved to SF thinking this is a short 3 month stint we'll finish YC and we'll head back but you know I've decided to stay back here so [laughter] that that's a bit >> yeah and I don't understand why I'd want myself to have a property in SF one day sooner or later it's such a a hub for technology AI and progress and growth uh amazing place although it has some problems as well for sure um my first question sort of odd if uh the Chad GPT moment was created by tiptop talent, right? Like top brains and consulting has a monopoly on top talent, right? Like it's [clears throat] not the first time I speak with top guys like you that have been at JP Morgan or McKenzie.

How do you reproduce that? How do you get like a monopoly on >> the the tiptop best young talent? Great question, Charles.

I think the the one thing that I've been like spending most of my time and have been obsessed is been with hiring. And essentially, if you think about kind of what is needed, it's agency, right? And and I think that's an overused term, but that's kind of what we look for.

most I mean like for the lack of a better word as you're framing it right top talent most of these people are really just ultra high agency you kind of throw them in any situation they just figure out uh what needs to be done um and I think that's just a function of a lot of things ranging from you know your childhood upbringing to kind of where you ended up being so I think at this point that's probably the number one thing that we look for it's less pedigree it's just agency we've hired a lot of really like cracked engineers who do not have like let's say the JP Morgan or the Mackenzie pedigree, right? But ultra high agency where you you just know that they're kind of delivering power. So so so that's kind of at least how how I see the world to be.

Um and yes, you do happen to have a lot more of these in in some of these elite institutions which also I think now it's changing. I think like you go back in the 90s 2000s every B school like from Warton to Stanford they probably all wanted to be in some consulting firm. We are right next to Stanford with a new office and you know often speak to folks there both from the business school as well as undergrad.

Um YC is the new fat there right everyone wants to sort of work for a startup or YC company. So we seeing some of that happen in the Bay Area, but uh I think it's it's it's the agency piece, >> right? And if I would ask myself the same question, um I think these guys have been well, first they're very valuable companies, right?

Like JP Morgan, Mckenzie. So they can pretty much afford uh anything and everything. They have a gigantic brand.

They have a good onboarding program as well. Uh, I think I think that's how you sort of reproduce it. Tough to get a mode there, but seems that they don't have a moat on technical talent, right?

Like you hired like the first best young engineers in your opinion. Is there like a flagship company? >> Yeah, I think it's changed a lot, Charles, since Chad GPD, right?

Like I'll probably nail that November 2022. >> Open AI that gets the best, right? Yeah, exactly.

So I think I think >> yeah, so if you kind of draw the timeline chart 2022 when open at least charge GPT I think world just changed in terms of talent. But if you go back a little bit right like maybe postco [snorts] um I know a lot of really uh good uh tech talent that was in Mckenzie for example right like that I worked with so Mckenzie has an AIM called QB uh but even there uh a lot of my friends that I worked with were like hey you know I I'll I'll move to X or Meta I think Fang you know as you would call it probably were were the with favorites probably everyone wanted to kind of be in one of those super niche teams within these large tech organizations. Um, and I think they probably got all of the tech talent anyway.

Uh, but I think that just changed drastically with OpenAI. And now I think every every AI company on the block here in SF is is probably the one where uh I think it's both, right? Demand and supply.

Everyone wants to kind of be doing something at the cutting edge. Um and and the other thing adding to all of this is also with coding agents like you probably don't need as many developers or engineers. So a lot of these big companies now are also rethinking what their talent pool looks like.

Um I have a friend who works at AWS and I think they were telling us how uh the whole bedrock thing right now the entire stack is probably being rewritten by a handful of engineers, right? which again uh not surprising but that then suddenly becomes ultra challenging for for everyone else >> and let's take that young software engineer um >> let's give them three choice right because you said uh they they're they're high on agentism or whatever that means why not start their company first and second why choose your company like a YC company over a meta >> yeah great question I starting so everyone we look for to get as an archetype is eventually they all want to be a founder so I think that's a really good trait to have where you want to build something but I do feel that based on sort of the industry you want to target having some sort of experience helps if you're trying to do a consumer app like a Facebook or or like a perplexity I think uh it's it's okay not to have any experience right because you're building for the consumer world so it's okay to kind of just go throughout experience and that's where you have these breakout stars 16 year old founders and things like that. Um so in our world we are very specifically focused on enterprise and within enterprise it's pharma and life sciences.

So it's a very niche industry. Um so you people will not take you seriously unless you can speak their language. So if you want to focus on like hey I want to go work in banking or kind of disrupt how um the insurance industry works you need to first understand what sort of the ground reality is.

So I always encourage people to kind of spend some time in the industries uh to to just figure out right what you want to do. So I think that becomes important. Um having said that now I think coming back to sort of the incumbents versus startups right I feel we are and I think everyone says this at an inflection point now and I think that's very real right so if you look at all the SAS companies probably in the last couple months they're all down by 20 30% and in the last 6 7 months Salesforce most of the CRM right like down 40 50%.

And the reason that's happening is because I think most people on the wall street and everyone else realizes how threatening AI is to incumbents and that makes it a very unique opportunity to completely rethink things from ground up. >> [snorts] >> Having been in a lot of these large organizations, I can tell you that, you know, when there's a threat, you know, it's really hard to say guys, let's completely rethink, let's cannibalize ourselves and like reimagine and rebuild, which while the intent might be there for a few years, that's just not happening. It's a lot easier to just build something from scratch.

Um so I think that's why anyone who now is serious about kind of building something and you know creating impact uh just based on probabilistic terms the odds of doing that with a startup are so much higher uh that it just makes absolute like mathematical sense to be working with a smaller company. >> Right. And tell us real quick what you're building here because it seems yeah obviously you're targeting life science.

You have a bunch of products slashfeatures, right? Like tell us how you're different also from your competitors. >> Yeah.

So, think of us as sort of this agentic and voice AI platform for life sciences go to market team. So, most pharma companies have just two verticals, right? So, one is R&D.

So, I'll kind of bucket them into two things. One is you discover the drug and do a lot of research around it. But once you have the approvals, you want to put it out in market.

uh that's where we come in. It's on the commercial side of the house. So pharma's end customers are physicians and patients, right?

So how can you completely reimagine what engagement with this group looks like? So we have a suite of products to support this group. So yeah, so so Jarvis for example is one of our flagship products which is an agentic and voice AI product for farmer go to market and field teams.

So if you're a farmer rep that's going and visiting a physician uh and typically you know you you're like hey who should I talk to next uh what are their top priorities what should I talk to them like what are the questions they might ask me so these are all the things you have historically all of this data existed in different silo databases maybe some of it sat within your CRM some of it within some AWS S3 bucket right so Jarvis today sits on top of all of that has context across everything so you and just talk to it. Uh yeah, and we shamelessly chose the name Jarvis from Iron Man, right? So it exactly does what uh Jarvis does for Iron Man.

Um >> yeah, >> so that's Jarvis. Beyond that, right, I think the way we also think about it is an AI that can actually be the rep for physicians. So that's third, right?

Which is can we have an AI rep that can directly engage with these end customers. You're already seeing some flavors of it in other industries, right? like the AISDRs, AIBDRs, but here you want them to be very specific uh and and a large part of our focus is compliance regulations because those become really important in our world.

>> Got it. And um I mean this is for reps, right? It um I think it's like yeah the least interesting feature in my opinion.

The most interesting one is like the the one in which you're probably helping accelerate um studies, right? like uh clinical studies. Uh this is I believe it's called Ader or Ether.

>> Yeah. So this one how how advanced is it like your most popular feature must be yes Jarvis it's for sales sort of right but >> uh do you believe that either will um outpace the that feature that >> yeah so I think yeah so I think that's sort of the eventual goal where so today it's all mostly human driven right uh because it's a very compliant uh and regulated industry it's still very human-driven uh we do think there's a world where it'll be hybrid, right, with both humans and AI agents. So that's kind of where Jarvis and Aether will coexist together.

And of course, I think over time there will be a completely new paradigm where maybe the go-to market model is all just AIdriven, right? And maybe there could be some very specific teams and brands that want to have an AI first way of engaging their end customers. That's where ather will become really powerful where you just have an AI native way of doing customer engagement.

>> And isn't it frustrating sometimes to deal in an industry that's so regulated like sales cycles are are pretty long and it's always like an eternity to make deals happen. >> Yeah, tell me about it. No, I think uh I think that's that's that was an intentional choice.

So I think it's both good and bad, right? So good in the sense the barrier to entry then becomes significantly uh higher as well because unless you very specifically know how the industry works and all the quirks of it, you can't just have a new startup pop up and try to take over. But at the same time, yeah, the challenges are a lot of these uh companies are still very heavily guarded by incumbents, right?

Including the consulting firms, including massive organizations. So they'll always have u a lot more stickiness. So in fact uh that that's an interesting challenge just for us which is very unique is that we often go up against incumbents and very few startups that try to do anything in the space.

Um, so that's that's but I would say that's an interesting problem to have and and you know we are really excited when we go into these situations where we are competing with someone who's been around for 1500 years. >> And talking about the incumbents, you feel um McKenzie will survive for example and all these big corporations they do have piles of cash but they're so slow in my opinion and bureaucratic and I think we've seen that cash is is not necessarily a moat. Maybe data can be but even data they're pretty slow to make use of that data.

Do you feel that the younger YC company have have a major edge in this day and age? Can they outpace these these giants rapidly? [sighs and gasps] >> I think so.

I think at a high level yes u I think you you definitely have the advantage if you sort of you know uniquely understand the industry and the problem you're going after. But having said that right, I think uh you know something like a McKenzie right like inherently your business model is to be kind of reinventing yourself and just be up to date. uh you know I often joke with uh with my colleagues and our customers saying hey tomorrow if there's a spaceship that lands uh on earth and aliens come in here McKenzie will have a practice line saying hey we can help you reimagine your business in a world where aliens are there right uh and I think that serves them well um just to kind of be at the new thing but having said that yes there is a significant challenge you kind of need to keep up with the pace which is really hard um unless you deeply understand kind of how the technology is moving and you have the right set of people, >> right?

Um with vibe coding for example, I think we can bring product to life incredibly quickly. And then if you master GDM that they still have a moat also that we don't have, which is trust, right? Um not only brand but just like building relationships with customers and being so stable and humans don't take a lot of risk, right?

So that's always hard to to get over with. I do I do this podcast system to >> to to get over that. But what do you have to say about like vibe coding?

I I've seen you often talk about this and the most importantly the open claw moment which is the proof that one [ __ ] guy can outpace open AI and tropic and and all of these guys with very few resource but just like massive willpower and massive knowhow I would say. >> Yeah. No, I think all of these are very real and I think VIP coding right is also very real.

Uh but I think it's just a function of right like call it the space you're in. As I said if you're trying to build a consumer app or a tool where you can just help individuals let's say build kind of their own business of any equal to one. Hey, I'm I'm a single solopreneur.

I want to build some cool apps, put it on app store and maybe make some money as a side hustle. I think that's great. But in our world, we are in the enterprise play.

I think the most important thing in any enterprise sort of scenario is trust. I think that's kind of what you are getting to, right? [snorts] Uh if they don't trust you, you know, it doesn't matter if you're uh right like open air or or anything else.

So I think a large part of our work then becomes making sure we build that trust and of course trust takes time and it's it builds over time and that's where we often see that when we let's say work with a certain client for the first time they then are like oh we like you guys you did a great job here and kind of hey let's rethink and let's rebrite like maybe more things and it kind of um compounds from there but again that's a enterprise specific nuance I would say versus if you're just trying to build something for the general masses no one cares right uh if if you give something that works and if it's cheap you'll have adoption and I think that's the difference I would say so I I do see wipe code working really well in and and that's kind of what we are also seeing right these are mostly all consumer style products versus enterprisegrade applications um there I think trust becomes a lot more important >> what do you have to say about the slowness. Well, let me rephrase this question. What is wrong with our current system and the FDA?

Right? You've probably heard >> comment on the the FDA, which very much is in line with uh what I think about them. I think they're trying to protect people, but in the end, they're actually protecting them from the cure.

So, they're more they're killing more people than they're helping in my opinion because they're taking so slow to approve drugs that should be in the market [ __ ] now. and saving lives. Um, and and not just that, but like approving technology and all the compliance [ __ ] and just dissuading entrepreneurs to to get in that industry and launch new stuff and try new things.

It's it's very intimidating to to get in the the life science. So, >> how could we improve that? >> No, I think that's a great point, Sal.

So, in fact, earlier this year, there was the JP Morgan Health Conference in San Francisco, right? And that's kind of where everyone from the industry was there and I was able to attend it and we had the chief of the FDA uh who had a session one of the keynotes and I [snorts] think two things stood out there from from that right one was this idea of hey we will not add processes just for the heck of it so I saw some intent there to kind of accelerate things. So I think he spoke about basian statistics right some way of if we can kind of statistically look at similar things in the past and if it makes sense we'll probably not add more steps to the process and [snorts] I think there was also this general intent to say hey AI is real AI is good if it can make things faster we do want to adopt it I think we are waiting to see some of those happen in action right so for example we build AI agents that can directly support physicians and there are these painfully uh n number of things that just don't make sense but they're there as rules of hey you need to always show like this massive uh thing or right like no one's like that but but I do think there's intent so I think I think that that will happen and the second trend I think Charles that is really interesting here that's happening in the space is I have a [snorts] few friends that had AI companies that were trying to do drug discovery right right selling into pharma to help them accelerate drug uh discovery of drugs [snorts] Some of them were just uh so fed up with the fact that it's taking so long they kind of reinvented themselves to be the pharma company.

So they were like instead of selling software to find drugs faster I'll use my software find my drugs and put it in clinical trial. So a lot of them now are you know these AI first pharma companies. So I think a lot of that will happen as well where these startups instead of trying to sell softwares that help big pharma make drugs will just be like screw it I'll kind of make my own drug and put it into trial.

Um so I think that's an interesting trend to to see as well how that pans out. >> And you know what insight I had just listening to you is that big pharma is probably a product of why of of uh why the FDA is is that way. They create such like high barriers that they actually create just a few couple of winners that have a nons surmountable mode because these guys invest billions in clinical trials and hopefully that will be uh changing with AI.

Um one last question, you raised 5 mil I believe a couple of months ago. Uh congrats. What's the the next step here?

What uh what do you have in mind for this year? Yeah. So we've already over the last six months Charles scaled our team to 5x.

So we have a massive team here and we are split between Bay Area and few folks in Bangalore, India. So the goal is to kind of continue growing our our team [snorts] and then given the growth trajectory we are on, we'll probably soon be doing uh our subsequent round sometime later this year. So that's the hope kind of get more customers, get more revenue, the classic.

There you go. Well, thanks for being on. Where can uh people find out more about you, Supre?

[snorts] >> Yes, I think LinkedIn is the best place. So, hit me up there. But wonderful being here and thanks for inviting.

I was supre.