Loading summary
Transcript42 lines
- [00:01]
A
We did the first ever fully autonomous patient interaction powered entirely by AI.
- [00:08]
B
Welcome back to another episode of Builders. As always, this show is brought to you by Frontlines IO, Silicon Valley's leading B2B podcast production studio. If you're bringing technology to market and want to learn from your peers, we have a library of more than 1200 interviews with Venture backed founders and marketers. Where they talk, all things go to market. Of course, if you want to launch your own podcast, we offer podcasts as a service to more than 80 tech startups. The idea there is very simple. You show up and host and we do everything else. Now with all that said, let's jump into today's episode. Our guest today is Jamal Soni, CEO and co founder of Insight Health. Jamal, welcome to the show.
- [00:47]
A
Yeah, appreciate having me. Nice to meet you.
- [00:49]
B
Of course. Looking forward to the conversation. So I see that you spent seven years at Segment. I'm sure that was a ride. What did you learn from your time there? Oh, wow.
- [00:57]
A
The importance of go to market in any business. I mean, I knew it was important before, but I think one of the biggest things is Segment had an amazing product. Just to give some background there, it was a customer data platform. They connected data sources to data destinations. At its core it was an infrastructure play. But then as we moved up market, a lot of the marketing folks really started to gravitate towards the products we're building. And we eventually evolved the business towards more of a marketing buyer. Very much thought it was a developer first sales motion. Realize very quickly that can only get you so far. You have to bridge that gap between technical and marketing, slash business buyer. And I think go to markets would ultimately helped Segment achieve success. So in 2021 we got acquired by Twilio for 3.2 billion. And now it is a critical part of the Twilio offering.
- [01:44]
B
Is there a big lesson that you took away with from there? And I'm sure there's tons of lessons, but is there like one specific one that was just almost like burned into your brain as you approached founding this with your co founders?
- [01:54]
A
There are two things I think the first is the importance of discovery at every part of the deal. So early stage, really understanding the customer needs making sure that you spend a ton of time asking the right questions and that doesn't stop once the deal is closed. We often work post sales trying to make sure we can continue to understand, you know, maybe the economic buyer or someone has changed or maybe the scope has changed from the customer perspective. So you've got to always Be discovering is what we would always say. And then the second is the importance of sales engineering. So when I first joined Segment, I started and led the sales engineering function there and did that for many years. We were moving back into product and the importance of sales engineering, I think I cannot overstate that. And you know, not to discount the amazing salespeople that the account executive and CFMs that we have, but sales engineering, it is this very unique set of skills that we had to hire for which were somebody you can throw in front of a challenging business buyer that could speak the technical language and can understand the mechanics of the deal to help push it forward. And they're almost as Swiss army knives that any sort of deal, they kind of figure out what the next action needs to be and they help push it forward. So those are the two things I would say that are cemented at the back of my head.
- [03:05]
B
And when it came to the founding of Insight Health, I'm sure that if you're like most, you had a list somewhere, whether that's digital or written down, of things, things to go after, problems to solve for Insight Health, maybe just intro the problem that you solve and explain why you decided that was worth going out to solve.
- [03:21]
A
So when we first started Insight, one of the big insights we had and I would actually give the credit to the two doctors a bit about our team. Myself and Saran are more engineers at heart. I led product and go to market for a while at Segment and then became Twilio. We started the healthcare team there. Siron prior to Insight, led infra on security at Twilio and then before that at visa. We have two physicians on our founding team. We have a neurosurgeon, Dr. Pankaj Kaurer, and a EP cardiologist, Dr. Eric Stecker. And a lot of the credit goes to them. So the story is actually on a run. They were thinking about the opportunities of AI in healthcare and the obvious one even when we started was AI scribing. And I'm sure we'll get into this, but I have a separate story about AI scribes. But when you think about the problems that we really feel are the most impactful, it's. It's how do you make sure that each provider has their own physician's assistant that can capture information from the patient, really help with the history capture is what I came to understand very clearly now is that up to half of a first visit is that history capture part of it. And that really sets the trajectory of that patient provider relationship. So the problem that we set out to solve initially was to build this physician's assistant powered entirely by AI to meet with patients ahead of time, do things like history capture, potentially meet patients after the visit as well to help with follow ups and some of that routine touch points. For chronic disease management, we did the first ever fully autonomous patient interaction powered entirely by AI. It was for an oncology practice on the west coast. The first one was actually chaperoned by a human to make sure it stayed on the rails. And since then we've done probably about just over 4 million encounters now. So in general that was the problem we solved. And we've since expanded of course beyond that. But that's what we started with.
- [05:03]
B
When did you get your first early customer? How long did that take?
- [05:06]
A
So we started at the, probably building mid 2023. I would say our early customers were more design partners. So actually one of the first ones was that oncology practice and that was probably three to four months after starting the company. So our first customer, so to speak, was this oncology practice in Portland, Oregon. We started with intake and are now expanding to various other products. The first paid customer was probably January of 2024, maybe four months later after starting the company.
- [05:39]
B
That's really fast for healthcare though, right?
- [05:41]
A
It is. The healthcare procurement cycles are insanely long. Part of what we decided in the early days when we first started was to go after mid market. So mid market for us is think like seven providers, seven doctors basically at the small end and up to about 50 to 60 on the high end. Those tend to move a little faster. We're talking about the one end of the spectrum, like the kaisers and the HCAs of the world. They're very large. They take 12, 18 months to close. But mid market can close within three to four months, even less sometimes if they have a burning need.
- [06:13]
B
Do you think that'll make it difficult if you ever move upmarket and go enterprise?
- [06:16]
A
We have already started doing that. Part of what we spent a lot of time doing is building the foundation, building the credibility, building the machine that can help support the long term growth. The best quote I've ever heard is you never want to go out fishing for a whale in a dinghy. And I think the reality is if we had gone straight to enterprise when we first started the company, we were just too lean. There was just no way we could do it. Now that we're, we're just over 20 people now. We're continuing to scale, be at 25 by the end of the summer. We have post sales we have for deployed engineers. We have resources that we can deploy to help make these larger enterprises successful. So we are moving up market. I think to go back to your question, early days we were too lean. Now we have the machine to actually successfully go approach enterprise.
- [07:02]
B
If I look at patterns, trends from the conversations I have with founders, forward deployed engineers seems to be one of those hot trends. It's hard to say I can get away from a conversation without that coming up at least once. Now what have you learned about building out that discipline in the org so far?
- [07:17]
A
I think the reality is it's not a novel concept. I mean the term might be new, but Google back in the day had the notion of a customer engineer. Arguably even the solution engineering function that I had the opportunity to start and run at segment is very similar in the world of AI. The beauty of it is we can actually create a model where we're incentivized by outcomes. Not necessarily like billing people by seats or even on the customer side paying by number of people using the product. You can actually procure software and tie success to real outcomes. And we try to do that through all of our products. So an outcome might be more the non clinical side might be scheduling a visit on a clinical side, it might be doing a full history capture and getting a patient schedule for a procedure. And the reality is that spans pre and post sales and it's not just a technical track and not just a business track. It really has to look at the entire success of what the solution that you're offering is. So when we think about forward deployed engineers, it is very much that skillset that I mentioned before. It's somebody who has the technical understanding, has the ability to understand the business context and can help sell when needed because the selling doesn't stop once the contract is fine. In fact, that's really why it starts because you have to deploy, make sure they're successful and then continue to help them be successful. So the deal on our side expands and continues to grow.
- [08:36]
B
What advice would you have to a founder listening in who's in the early days of standing up this program?
- [08:40]
A
I would look to people that have done this successfully. Obviously the one that comes to mind as Palantir, I think they coined the term but find ones that have been successful in your particular industry. So you know, in our case, I mean I'd look at the ethics of the world, right? Like EPIC has this function where it might be multiple people but the function is still the same. You have somebody that during the onboarding process of epic, which is this massive PC software, those functions are taken care of. So we, in an early days, we talked to a lot of people that either worked at Epic or had interfaced with them, even on the customer side. So clinicians and hospital administrators that have worked with EPIC to understand what worked and what didn't. But yeah, I would look to people who are doing it really well and just copy what they've done because they've spent the time figuring it out. I think the other side of it is like, really narrowing down, like, because it is a very wide scope. So figure out the things that you want to really, really index more heavily on. And what we found, even at segment, we're hiring solutions engineers. Usually, if you think about it as the Venn diagram, you have sales engineering and maybe like a product mindset, we're thinking about the business track. The intersection of all three is very much a unicorn profile. You're usually going to get someone that indexes heavily on maybe two of the three. So just figure out, like, what you're willing to coach them on or find areas where you think maybe it's a little more important. So for us, those two things are certainly the technical side of it is super important. And then sales for us.
- [10:03]
B
Right.
- [10:03]
A
I think the product mindset, we're fortunate to have the positions on the founding team. We have a lot of folks that truly understand how the two meet together. But I think for us, the two of the three we tend to index more heavily on were really that sales side and the technical side.
- [10:16]
B
This show is brought to you by Frontlines Media, a podcast production studio that helps B2B founders launch, manage and grow their own podcast. Now, if you're a founder, you may be thinking, I don't have time to host a podcast. I've got a company to build. Well, that's exactly what we built our service to do. You show up and host, and we handle literally everything else. To set up a call to discuss launching your own podcast, visit Frontlines I.O. podcast. Now back to today's episode. Something else you mentioned earlier was, you know, building out that foundation, building trust. I would guess having two positions on the team has to be an important layer there that's added a lot of trust and credibility. What else have you done to really move the needle from a trust perspective, from a credibility perspective?
- [10:54]
A
When Saran and I started the healthcare team at Segnitulio, we spent a lot of time digging into more of the technical. Like when we say HIPAA compliance, what does that actually mean? It's the Technical infrastructure that can support that. We're also SOC 2 accredited and there's various components to that. And as we continue to scale, things like high trust and other areas are even higher bars to hold ourselves to. What it boils down to is really having a solid technical foundation to help us scale. And Saran, to me, to his credit, he's done a lot of that at companies much larger than ours, like Segment Twilio. We've even at Visa before that. So really making sure that you one build that function. So one product of that just to kind of share a tactical example, in the early days and I think even to this date there wasn't a evals framework that would work for our use case. So for example, we're having a conversation with the patients in real time. How do we make sure that things are staying on the rails, that the models are performing the way they should? And then near real time, how do we make sure just after the conversation, should there be any red flags or should there be anything that needs to be escalated? We can do that right away. There wasn't a framework that would work. We actually built our own. We call it Safe AI and it has the ability to do that both real time and then near real time checks. So things like that are super important. That helped from a credibility standpoint because in any meaningful deal, you're going to get an IT person, you're going to get somebody who deeply understands the tech and will look for stuff like that. Like you know, of course they're choosing you as a partner and whenever personal information and especially personal health information is involved, I think there's always a higher bar that you held to.
- [12:32]
B
That's good to hear. That's what I want to hear about healthcare and data. Now, when it comes to the two positions we just talked about on the team, are they, I don't want to say this in a rude way, like are they actual co founders or are they just advisors that are advising you and supporting you?
- [12:45]
A
I love this question for so many reasons. I think that you cannot build a company in healthcare without having a deep understanding of the clinical workflows. And we talked about trust and credibility. Without that, I mean you're basically hosed. So for us, both Punkage and Eric are involved in every decision that we make. They are just as much founders as Saran and myself. And I think to that end it has allowed us to move as fast as we can possibly move in an industry that seems to be changing every day. So Saran and I come from more of a technical background. And Eric and Pankaj have this whole other set of experience that is so different than what we have experienced and so relevant to I think the startup ecosystem. So just to give some background, Punkage is a neurosurgeon and Eric is an EP cardiologist. Each of them practiced for 15 plus years in their respective fields. And I think the stuff that they're thrown at a procedural list, so the stuff that they see on a day to day basis, you know, not to discount the clinical side, but you never know what you're going to get when they're working with patients. It's just like such a amazing thing to see. And I think in the startup world in that sense, I think the uncertainty is certainly similar. So to that end, I think they've been such a critical part in helping us move through challenges. Whether it's a business challenge or whether it's even like a product or technical question. Having their expertise has been instrumental to the success of the company.
- [14:05]
B
Are there any insights you learn from a go to market perspective from them? Like I would imagine they're physicians, they're probably not the end buyers, weren't the end buyers of technology. But was there anything interesting that you learned from them? Just about how to market and sell healthcare like this or healthcare technology like this?
- [14:21]
A
Yeah. And it's interesting because they both approach it from different perspectives. So Eric is at an academic system and Pankaj is part of a private practice. And from the enterprise side, just learning how enterprises procure software at that scale has been, you know, I think having a clinician perspective, you know, he's obviously part of some of these conversations in terms of procurement and then punkish from a private practice, you know, they have the partnership, so each provider is part owner of the practice. So obviously he ended up stepping down as one of the board of directors just given the conflict. But in general, given his previous experience, he was on the other side of the table procuring products and things like that. So we learned a lot about just the way these organizations make decisions. The one thing, the insight that we had early is that you really need three different champions for a deal to win. One is you need a clinical champion. So that might be the chief medical officer, it might be a single clinical champion that's carrying the torch. You need an administrative champion, someone who's directly responsible for day to day operations. And then you need some sort of, you don't necessarily need a champion but you need some sort of like economic buyer, the person who's Going to sign off on paying for this and two of them might be the same person. We learned that early just from the experience that Eric and Punkage have had in their respective lives procuring software and just being part of those buying cycles.
- [15:40]
B
As you gear up for this move up market, or I guess I should say that the move up market is already underway. As you're making that move, what's top of mind for you? What do you think that you as a leader and as an organization have to really just get right to really nail.
- [15:53]
A
There's a few things we talked a little bit earlier about building the machine to really support that growth. So hiring sure out of the gate is the thing we have to get right. And we, you know, give you example. We now have three post sales clinic success managers and each of them are amazing. Part of that is making sure that they balance the success of the clinic versus just understanding like the evolving needs of healthcare organizations and making sure they can interface with product and help improve our product over time. So hiring for that skillset has been. It's hard to find those roles but we've been very, very, very fortunate to find three amazing csm. The other side I think is figuring out what not to do. I think in the world of AI, one of the challenges with this forward deployed model is that in many ways that balance between what is product versus what is services becomes a pretty big challenge for a lot of companies. And I have a firm belief that if you're selling services, you can't really book that as annual revenue because it could just be somewhat ephemeral and tied to implementation and things like that. I have to imagine it's not unique to us, it's unique to any business that's operating in AI where in a world where you're selling the outcomes, how much of that is the actual product that's recurring revenue versus how much is actual implementation or just stuff that you have to carve out of the revenue anyway. So that's the other thing we really have to get right and make sure we don't, you know, I would say pry ourselves trying to do everything for everybody and not actually have a core product. I think we've done a really good job done with that to date and find the common patterns and continue to evolve the product. But it's something that we have to do to a probably a greater degree when the contract sizes are larger, when the organizations are more demanding, things like that. This show is brought to you by the global Talent company, a marketing leader's best friend in these times of budget cuts and efficient growth, we help marketing leaders find, hire, vet and manage amazing marketing talent for 50 to 70% less than their US and European counterparts. To book a free consultation, visit globaltalent.co.
- [17:52]
B
when it comes to the competitive landscape, how do you kind of break that out? And you don't have to tell us names, but if you just look at this market map, if you think about your buyers and they're evaluating solutions, what are the general buckets that they're evaluating on?
- [18:05]
A
So the first thing we see is most organizations look to their ehr, the Electronic Health Record, first, to see what they have. And I said, we come back to the scribe story. So the early days, we knew the scribe market was very commoditized. And for those who don't know, AI scribes are very much like those note takers we have before providers that can document and help with charting. We thought about that market and basically decided to build because we got asked by our customers. We built an AI scribe and we offered it for free because that ultimately is where the market's going to converge to. And turned out to be true. Like most EHR companies now have them, the price point is dropped significantly. What that taught us though, is that the winning solution are the ones that are natively offered by the EHRs. And we see that in the bigger enterprise. We see Epic has their own. We see that Athena, that's an ehr, they're starting to spin their own. We actually partner with EHR company so that they can offer their own. So it actually became more of a channel for us to sell through. So using that as really a example for how this market could unfold, there are parts of our offering which we think are germane to what an EHR offering should be, and there are parts where we're confident that the EHRs are not going to play. So that's how we carve up the market. We look at what is really a feature on the EHR versus what is a real product and differentiated product at that. Some of the areas we look at when we think about, like Intake, for example, as the first product that we built out, like we're meeting with the patient, we're doing the history capture. There's a lot of clinical reasoning that we built in were operating at the level, or in some cases, you know, higher than physician assistant level, just with all the training that we've done over the years. To that end, you're really replacing a clinical encounter and from the patient perspective, just to share some context There we're not replacing the patient visit with a robot. We're actually just replacing the parts where, you know, I'm sure you felt this way, I've certainly felt this way myself, where you get asked the same questions over and over again and seems like when you see the doctor, you get asked the same questions again. We're replacing that part and then providing the physician with really a quick summary. So that in that example, that is a clinical encounter that would sit on top of an EHR and that's how we see the market. So intake is certainly part of it. Moving upstream, we've gotten pulled into referrals. Referrals is another great example where there's parts of it that touch the ehr. There's parts of it that touch aspects outside of the ehr, like maybe a portal for a payer or engaging with the patient to fill in any gaps. So there are parts of that workflow that ultimately will sit outside the ehr and then after the visit as well, the routine follow ups, checking in with the patient, things like that. So all that to be said, I think that's how we think about the competitive landscape. In terms of the first competitor you really have to clear is the ehr. And then in every single, I would say slice of the clinical interaction, there are going to be some strong players. The way we try to tackle it is we really see it as not just a point solution. We're really end to end the entire ecosystem around that specialty provider, from first touchpoint through to ongoing care. We try to automate as much as possible with that routine work.
- [21:13]
B
And a final couple of questions here. I know that we're up on time. If you just were to zoom out, how would you describe the state of AI in healthcare right now? What are buyers thinking? Are they at like the point of fatigue? Is it just deep urgency to go and implement solutions? Like what's the high level sentiment?
- [21:28]
A
There's a lot of optimism. So I'll start with maybe the clinicians and speaking of punk and generic themselves on our team, but also just talking to our customers, there is a lot of optimism because what has happened over the last few decades is that clinicians have gone from actually delivering care to being mostly like data entry specialists, which in a fortunate reality from the reimbursement structure we have here in the United States to just the general flow of things. And I think that optimism has come because, you know, I honestly credit the AI scribes for helping alleviate that. I think the jury is out in terms of what the monetary ROI is like number of patients seen or what have you. But I think there is zero doubt that it's helped work life balance for clinicians. So because of that I think AI has earned a lot of credibility and there's really an opening for other new avenues. And part of it, like us, for example, like the intake side, the customers that we have that use that product are in many cases they can't see themselves practicing without it. So there's a lot of that in the clinician side. Just like optimism and I would say proven value from the administrative and more the OP side. I mean, healthcare is somewhat unique in the sense that every year reimbursements might stay flat or in some cases go down, but your costs go up. And in a world like here, you know, just around the corner from where I live, there's a hospital that closed down couple months ago because they just couldn't make the numbers work. And I think we see that, we hear that story over and over again. So from the administrative and ops perspective, certainly optimism, I think it's cautious optimism because they've been oversold for years on things that will help save costs and things like that. But I think this is the first time that we've seen adoption from mid to large size practices at this rate. So certainly there's a burning need, certainly there's an opportunity. And I think based on what we've seen, there's a huge pull from the market.
- [23:20]
B
And final question for you, let's maybe just zoom out here and let's go 10 years out. So what's the big picture vision 10 years from Dave? What's the company look like?
- [23:27]
A
The way we think about healthcare today is that of course access is a big problem. The average time to see a specialty provider is in this country, I think six to seven weeks, sometimes longer, depending on your insurance, depending on where you live. And the way to bridge that gap is to help clinicians automate as much as possible of the routine work so that they can see 10x the number of patients. So looking at 10 years from now, we are moving towards building the AI doctor for specialty care. And the goal is not to replace the time you have with the physician, but it's to offload, really the parts that can be offloaded to make sure that the time you do have is much more meaningful and that will save lives. No doubt that will help clinicians see way more patients than they have the capacity to do so, and do so in a way that they can if they want to stay private. That's One of the challenges. So, looking out 10 years, we see a world where the AI doctor can handle the majority of the cases and monitor proactively rather than reactively and escalate to expert physicians where and when needed, who they themselves can leverage the latest and greatest technology to help diagnose diseases. Even the more esoteric ones diagnose them early. So that shift really goes from what is today a very reactive model to in the future, this proactive system of intelligent action, as we call it internally, that can reach out to patients automatically and really be this autopilot to help drive better health across large populations.
- [25:03]
B
And as a consumer, that makes me happy to hear because I feel like there's been so much money poured in here to healthcare, and I feel like my experience has stayed the exact same, which is not great overall, I would say, for most patients and most consumers. So you've ended this on a high note. I'm now optimistic about the future as well.
- [25:19]
A
I appreciate it. Yeah, we've never been more excited. It's never been a better time to build in healthcare.
- [25:24]
B
And where should we send people if they want to follow along for the next 10 years as you execute on this vision?
- [25:28]
A
Yeah. Check us out at InsightHealth AI. And we're also hiring, so check out our first page.
- [25:35]
B
Amazing. Thanks so much for taking the time.
- [25:36]
A
Appreciate it. Thanks so much, Brett.
- [25:41]
B
Well, that's all for today's episode of Build Builders, brought to you by the Frontlines. If you want more amazing content like this, visit Frontlines IO, where you'll find the library of more than 1500 interviews with founders, marketers, and other GTM leaders, where we unpack the tactical lessons from their journey. And of course, as always, if you do want to launch your own podcast, we'd love to have a conversation with you. Visit Frontlines IO Podcast as a service. Mention that you listen, mention you love the show, and we'll give you a 10% discount. Thanks for listening. We'll catch you on the next episode.