Welcome back everyone to the iCare Innovator series. My name is Rahan Ahmed and today I am so happy to be sitting down with Nittin Rai. He is the founder, president and CEO of First Insight Corporation. And this is the company behind Maximize, one of the longest standing EHR in practice management platforms in iCare.
And then we first met a couple years ago at the AI and iCare conference. It was just a pleasure getting to meet you. You're very warm. You ask a ton of great questions and you're also very generous with your time.
So I really appreciate having you on. I'm excited to be. Yeah, yeah. And just to sort of set the stage, I mean your life path is a little unusual.
You trained, you know, now AI is all over the place. But you trained as an AI engineer from what I understand. And you started First Insight back in 1993. I would love for you to talk about that story that you have about, you know, having $1000 in your bank and maybe your own interaction with the iCare field and how that you went down this path.
Sure. So just to give you context, I got exposed to AI in back in 1986. I was a, I think sophomore or junior in college. I went to Dalhousie University in Nova Scotia.
And I worked in the computer center as a consultant. So I would work with students from various departments. So I was in the computer center of the medical school, also in the business school. And as part of, you know, because I was a consultant, I get to meet a lot of PhDs and MBA students.
You know, they're all using sort of the bleeding edge of technology. And by the way, you know, Apple had just released the Macintosh in 1984 and they were just giving them away to all these universities and professors. So a lot of these students were getting in with computing at the onset. I mean, it was really because of the explosion, the personal computing market.
But back in the mid 80s, AI was a thing. And it was a thing in form of knowledge based systems and expert systems. And I just, you know, learn, I take an AI class, learning two languages, one called Prolog, which is logic based like programming languages, very suited to neural nets and Lisp, which was kind of the go to language for building expert systems. So having that exposure, I started supporting a couple of PhDs, including the professors in building expert systems.
I had exposure to something called a Lisp machine, which was a basically, you could build an expert system in this workstation and then using Prolog with a medical doctor who was, who was basically building a diagnostic system using Prolog. So that was my early exposure to AI in terms of building expert systems. Today we call use this loosely large, large language model. Right.
Really was designed for language processing and it has morphed into way more things. But the concept of the expert system was really rules based. And so that's the AI that I got exposed to was rules based. We studied neural nets, but they were too early.
Didn't have the compute power. Didn't have the data to make it fully effective. But anyway, after my junior year, I ended up in Silicon Valley to work for an AI startup, come big of quintus computer systems that provided the real language compiler, Prolog compiler, to build the knowledge based systems or expert systems recall in Prolog. And that's where I got real exposure to AI because my job was to commercialize a bunch of technology from Technion University in Israel.
Two PhDs that came from the AI field and they built this platform that plugged into quintus Prolog and allowed you to build visual applications. So you could design UX UI on workstations. And part of my job was to also build an expert system. So what I built was the system built in Prolog that could read data from disparate sources, connect to different databases and create computer graphics.
That's The end note, thank you for that sort of walk down memory lane a little bit because I think for a lot of people, they think AI is brand new that with Chachi, BT, that's actually the birth of AI. But in fact, you know, there's a long history, decades of AI work. The first AI and I actually used it, a lot of us would use it. The first conversational AI platform was a program called Eliza.
I can't remember the name of the professor, the PhD or Everett guy at some at MIT that built it in 1964. Yeah. You could talk to Eliza all day long. It was very rudimentary conversational AI.
So it's been around for a long time. Yeah. So and I think for bring up Eliza, that's an interesting concept. And that's sort of behind this idea that we've sort of talked about on our podcast with Scott called the Turing test.
And we're sort of like diverting. The Turing test, as you know, is sort of real test for AI about having a computer that's able to trick you into thinking that it's a human in a conversation. Is it true AI? Setting that aside for a second, I don't want to go down that path.
But in this sort of new generation of LLMs that everyone sort of is talking about, in your mind, is this an evolution that we sort of been on with AI? Or is this, do you feel like the past few years really has been a fundamental reshift in how we're talking about and doing AI? So if you look at the history of AI, it started with rules base. And then in the 90s, so as the internet started to take over, like e-commerce and as data started to get consolidated, as computing power got cheaper, machines got faster, we got into this thing called machine learning, which is also AI.
But machine learning is statistical, they're based, right? So it's still different than what's available in the large language model. But the largest language model was really, I mean, they were language models, but they were rudimentary. And they were, they didn't, you know, everybody was trying to solve the speech recognition problem, right?
A lot of effort into that. So drag and dictate, all these here, they're all somehow predicting, but not predicting, right? They're trying to pattern match, weren't predicting. So the prediction, which is probabilistic, happened with Google, publicizing their transformer algorithm.
That was the transformation, right? So that is the new AI. It's so funny because I met somebody at a conference, they were speaking and they were talking about AI. And I said, I'm from the, sort of the 80s and the 90s, and he goes, well, that was an AI.
I'm like, come on, it's, it was silly. It's just a different kind of AI. And I think, I think that was the transformation, was the transformer, which used vector math and sort of more of like predicting what the next thing is going to be, right? And that's how the, why it's called the large language model, because then they started to just suck everything in.
And, you know, we just invested in a company that worked very early on with OpenAI on their language model, because these are PhDs in language. Yeah. And so that was the foundational piece was to really allow the conversation. I mean, five years ago, AI was conversational AI, right?
Even pre-opening, it was conversation AI. I invested in a company, TrueArt, in 2018 or 17. It was conversational AI. They had their own language model.
So a lot of that technology wasn't using transformers, right? It was already a different kind of AI. So all that AI still exists, by the way, and it's so utilized, nobody is not utilizing machine learning anymore. There's still have pieces of that as compendiums or big thing or integrated.
It's not just everything is on in large language model, right? It's a combination. Yeah. And we try to spend a lot of time educating our listeners on our podcast.
And this is after all the AI and iCare podcast. So there is an iCare element to your story. So walk us through. So you're doing all this in the AI.
Yeah. So how does it get in? I can. Okay.
So my dad, actually, the way I ended up in Canada was my dad had a retinal detachment in 1974 and lost his one eye. The second I, he had to have some cryodone in India, and because he had lattice, but it was never explained that way, right? Because his prescription was like minus 13. Highest day, you know, hiding my opiate, right?
Thick glasses. I inherited some of that, by the way. So my dad had another accident, ended up, they screwed out his retinal surgeries in India. They lost power.
So he ended up in Nova Scotia because I had an uncle aunt there and they asked him to come over. He had to have surgeries and you know, I can't remember the name of this top retina guy in Canada and Toronto did the surgeries, saved about half his vision, but he had massive vitriol retraction. That was a problem, right? And so long term, the prognosis was not great that he would ever get his vision back, which he eventually lost all his vision.
I was, I was a pre-made in India and I wanted to be an ophthalmologist. And so I changed my mind and I said, I want to study computer science. And so my mom applied and I got into the Ohio University and undergrad program and very expensive, but you know, they did some sacrifices to my mom sold half her gold for me to pay for my very expensive education. And so I got an AIM, I ended up in Nova Scotia and I had three years of tuition.
So fourth year was was on me. And so then, you know, my fourth year senior year found this job at Traineeship, this AI company that, you know, kind of covered me for year four. And then I ended up in Oregon to work for a much larger company, but I had the startup boy, right? It's like, I always wanted to start my own company because my startup career in a startup in Silicon Valley.
So long story short, did one more startup in, in, in Seattle and realized I couldn't work for anybody because I just, you know, I was kind of an outlier. And I had seen the light in India. So what was going on in 92, I went to India, my company sent me, I had seen what was going on. I was wanting to do something in India.
So anyway, the combination of two led me to quit my job, moved to Oregon, started consulting with my previous employer, I had a lot of ideas around the internet and building some kind of an AI system, you ask a question and it would give you an answer. All right, and use the internet. But you know, again, I didn't have the money. So I went in from my eye exam because my, my optometrist a few years prior had discovered I had lattice and I outrofeed holes in my right.
So I was being followed by an optimologist, but I moved up to Seattle, hadn't done a follow up, go in his office, lock in, there is no paper chart. I go right into pretest, he's got a Macintosh, he's a Miss Craig Bowen, he still lives in his computer, he's got his own version. And I, and I clicked away my intake form, right, and he's got his voice in there too. So as soon as I click a stigmatism or click something, boom, his voice goes on and he's educating me about stuff.
Right. Wow, that's pretty impressive. Like, that's, some parts of the way we have that now, yeah. February of 1994.
Wow. Right. I just incorporated the company. And the first thing I did was go for my eye exam with me two years, right.
And so I, and then I go in his room, and he's got a slightly bigger Mac, sitting right to the side. And then there's this blue thing flashing on the screen. So he used a product called FileMaker, which is like this toy database, but very powerful. And I was familiar with it, highly customizable.
So I sit next to him, and I does his exam, and then he goes, and then he's like, click, click, click, and he had a 21 point form, by the way, which is a specific call. You're not, you're not optometrist. So you wouldn't know what I'm talking about. But it's a very cryptic form that the only the Pacific guys use, right.
So he programmed that form, and it looked like his form, by the way. And he's going in, and then he comes to that blue flashing, and he goes, did you see Dr. Seyke? I said, no, he goes, you need to see him, because you got lattice.
Remember? And I goes, that was flashing? And he goes, yeah, like it's reminding me. Wow.
Because I don't have any fallen notes from him in my, you know, records. I'm like, what the heck is this? Do you know, I said, what is this? So he explained, and I said, and he goes, he built it himself in the two years I hadn't seen him.
And I said, how many of you your brother or sisters use, you know, brethren use this kind of software? And he goes, nobody. And he was mostly right that nobody did. And I'm like, I got really interested, because I mean, I'm playing with a lot of ideas, right?
And this seemed like real, you know. And I said, do you mind if I just explore the market? And I just happen to know somebody who was involved in Acorn, which was a pharma company in Eichert. And I called him up and I said, yeah, like, you know, if you want to attend a show, there was the Academy of Ophthalmology coming up in a few months.
Anyway, I basically hired two interns from Portland State. We did a call campaign in Oregon, Washington, and we determined that nobody had it. But we also determined that the state of technology really sucked. Like they're all in like DOS and, you know, nobody was on Windows.
Windows was going to come out. Nobody was on a Mac. And so like, it just looked very clunky. And so I went back to him and said, Hey, I want to do a partnership.
And we went down the path and, you know, he decided he don't want to be in the business. He gave me the software. And I hired one of my friends, and did really used to work for it. And we pretty much we did an MVP before an MVP concept even existed.
So within literally within 60 days, I had a product because we took what he had and we determined he didn't have a lot of pieces and we built those. And that was the beauty of file maker. You can just really do what you can do today in scripting. And, and I sold, I basically, he took $3,000 for me, he said, just give me $3,000.
And I sold for three grand, I sold to his, his best friend who just retired last year after 32 years of using our software. And that was our first customer. And we were the first company to actually have a full end to end Windows based Mac based electronic health records software. And 30 years later, here you are.
There's been independent, no, never bought and sold. Did raise a lot of capital in 99 because we were going to do e-commerce. And that didn't work out too well for us. But we built one of the first, actually, we were the only system connected with DVI for lab integration and integrating with contact lenses.
I mean, we did this, stopped before anybody else did we were on the bleeding edge. We even built the first cloud system called minimize, but never released it. Because there was no demand for it in 2000. So we've always been kind of on the, on kind of the, the bleeding edge of the beast.
Yeah, I was going to say cutting edge where you're even very cutting, you know. And that's, and sticking with this theme of being on the bleeding edge. Now you're introducing a whole set of AI tools. Yeah.
And I won't call it bleeding edge because AI has already been validated. So this is the first time we're coming up with something where we waited. By the way, we've been looking at AI for the last the day charge GPD came out. I mean, we've been talking about machine learning and putting machine learning into our EHR.
We actually have a rules based AI system already built in coding inside of maximize.com. But we never called it AI because we're going to spook people out. Right. It actually does a little bit of the, the rules based stuff and has a little bit of machine learning, but not, not a lot.
They all make suggestions. But when, when chat GPD came out was sort of the, I like suddenly I just woke up to it. And, and then we started working on internal projects from an RCM standpoint, because we also own an RCM company called FASPA or a division. And we started developing using automation and AI internally for doing what we call today is called the billing assistant.
But we were using it internally inside of FASPA because very hard to scale that skill set. And most of that is out of India. So we started about four years ago working on it. And then, you know, evaluated, and then I started because I was making investments in AI.
So TrueLark was the commercial AI company that got into the dental space with a virtual assistant, front desk assistant type product. And at that point, we thought, okay, maybe we'll partner with them. Well, they didn't want to partner with us because they want to say in dental. Then we started looking at scribe products.
And we, you know, we started seeing like Suno and a bunch of companies pop up. So we actually went and did a complete end to end evaluation of every player in the industry to see if we could just partner versus build. And we realized they were incomplete. They don't understand eye care.
And there were pieces that were missing. So we decided to build our and that's Eva. So which we launched last last year when I first met you at the conference. So eva.ai is our multi-agientic platform.
I don't think anybody has that anybody else has the end to end. They may have pieces of it. And it's, you know, obviously a design specifically for eye care. But the way we put it together, it can actually be targeted.
It's template driven. So it can be targeted for other verticals. But right now we just really focus on ophthalmology and optometry and obviously also optical. So our vision with eva is to build as many agents as we can to really do things, reduce stock burnout and improve patient contact.
Because I would get pissed off going to my eye exams, especially as I started seeing ophthalmologists at Miami, University of Miami, you know, my go-go especially spent like literally 20 minutes doing this. Talking to me with his back towards me, I'm like, this is not acceptable, you know, and we were developing. They don't want to do it either. I mean, they don't.
They hate it. They want to be looking at you. Absolutely hate it. And especially the ophthalmologists, that's why they hire like expensive scribes because they don't want to deal with it.
Exactly. But then that's an expense, right? And so, so the eva scribe, I would tell you, is probably the best in the industry because it's really truly groundbreaking grounds up because it's completely customizable. And that's really the beauty of maximize was it's a highly customizable system.
We invented customizations, by the way, then everybody copied us. So you could tailor the system to your workflow and your form. Like, that's how we were selling maximize back in the day was, hey, you can customize to your workflow in your form because it's highly customizable. So eva scribe is the same way.
It's very customizable. And you can tailor the soul, it generates the soul, but you can tweak it. So it works with your workflow and your lingo and your how you do your exam. And the idea is that you just cut down the clicks to like two or three.
And you're just having a conversation with the patient. eva is recording it. Patient leaves. You can choose to work on it right then and there, or you can read it again.
But by golly, you're done by five. When you leave, you're done. You're not doing any more charting, right? And so does your pre-testing person, like, so it's the voice piece is completely integrated into the workflow.
So we will be the first full, fully functional voice, MR in the industry with the at that rate of which we're going with with some of the agents that we're launching. The whole idea is that it's touchless, right? Maybe a few clicks here and there, maybe on your mobile phone, maybe it's in your glasses. Oh, wow.
Yeah. Yeah. It's a, you know, I mean, that stuff isn't far away. But mostly it's going to be voice with very minimal screen time.
You know, we often talk about on our podcast about the hype cycle of AI, you know, and where are we on the hype cycle? And I think a couple of years ago, Scott and I will talk about, you know, how great things are going to be and, you know, all these great tools. And it's just, it felt like it took a little bit of time. And it seems like finally, as people are getting more comfortable using AI, I mean, a year is a is an infinite amount of time now.
It's like, and now we're getting much more comfortable about using different AI technologies. And so for the clinicians, because we have a number of I care providers who listen, what has been the number one or two things that people have found helpful with your solution? Is it I assume subscribing was super, super powerful, but what about driving is the killer? But it's the killer app.
It's the killer app because it is it's giving them so much time back. Both the technician and the provider, it's giving them time back. It's like, they don't have to do, you know, if you listen to some of the, you know, we put a lot of this stuff out there. It's also on our website, some of our customers, the early adopters that have been using it, it's just gives they leave when they leave, they're done.
They're not doing charting, right? They're done. And that, to me, is giving them time back, which is really the most important thing for them. It's just give me my time back.
And that's what they're, you know, the downstream efficiency is, you know, obviously all staff is going to get displaced. You still need your people. Maybe you leave one or two more people, but maybe the one's doing like a lot of the junk, the gunk work. Maybe you don't need, you need, you know what people need is that patients need face time, right?
Because when you have face time, you have quality of care. And we are losing that rapidly across the healthcare ecosystem everywhere you're losing that, you know, all the voicemails and the phone calls and the this and this and this, that front desk assistant cuts all of that. So front desk is happy, right? And so our, some of our, our KLA's, I think they're saying that their staff is actually happy.
They were early, they were scared. Hey, I'm going to lose my job. Now they're actually happy that they get their stuff done and they don't feel overloaded, you know? Yeah, we talk about that AI, ironically, is it may bring the humanity back to healthcare.
It's going back to the old ways, right? Always when they were paper charting, it was, yeah, they would write, but they wrote afterwards, you know, they didn't write while you were, you had actually would have a conversation with your doctor. And the problems of you write afterwards, you have to document all this stuff to like seven, eight o'clock at night. Or they may be recording it, or they just do a scribe, remember they used to just record it, right?
And then a scribe would enter it, right? Yeah. So that's, it's, it's like, it's kind of going back, it's going back to where things were, which is you want to have just, just, I think computing, and I'm, you know, partly, I'm, I'm an instigator of that, right? When you're introduced computers, you introduce another interface, and that interface requires a lot of clicking and data entry.
And yes, by entering the data, you're now getting more information, better information, you're, you're improving certain efficiency of communication. But you've also burdened people to do all the data entry. And that's, that's demanding, that's taxing, and it's complex for physicians. It's very complex.
I would say data entry of medical data is probably the most complex piece of data that you can enter, because you got to deal with all the coding and all this insurance related burden that had been put on people. There, you have to code things correctly to get the reimbursement. I mean, that billing agent is like 70% savings, 70% savings. And it's, and it changes and the rules change, and you have to hire staff and there's alterations.
Yeah. All of that. It does it. You know, the amount of time that people spend doing authorizations on the phone, we cut it down to zero.
I'm sure this is music to my ears. I'm sure a lot of the health care is incredible. Because, you know, we, because we ate our own dog food, we did it for our own benefit. And now the practices that are implementing it are benefiting.
Now, it's not going to do collections for you. So for that, we still have this fast-paced service to tweak and stuff like that. You know, we, we, we, you know, one version of billing agent, we're actually going to bundle some service so they can help with the AR part for people that want to do it that way. Yeah.
Because the collections is another nightmare, right? Just calling the insurance companies and trying to, you know, denial management and that stuff, you know, you can't fully automate that. You know, I mean, solve it. But the, the accounts receivable part is, you know, it's a, it's a, it's a real pain.
So by improving the frontline process, you can reduce it, but you can't completely eliminate it. So you still need people to do that part. No, yeah. AI is definitely going to touch every single piece, I think.
Well, the thing is insurance companies have instituted AI just F well. Yeah. So val, they're rejecting even more claims, by the way. So they reject the rejection rates is even higher now.
So I wonder if we're going to read some weird like game theory, steady state or something in the future. I don't know what they're doing, but you know, the whole point of the insurance companies not to pay. So I want to change, change gears a little bit here and talk a lot about your work. Because you also, you, you built the HR company, but you've also done a ton of investing with your elevate venture capital firm.
Talk to me a little bit sort of about that, how you got into it, and you know, what your fund is looking for and how you evaluate. And we have a lot of builders who are who listen to us as well and start with that. Yeah. And they help, that has a big, it's a big part of my interest as a VC.
So this was very random, by the way. So I never wanted to be a venture capitalist. It was not in my game plan, right? I want to, my goal was to go public and then retire.
And I tried retiring a couple times, and then I didn't. And the last time I tried to retire, I so the way the way this happened was in 2007, so I had bought my VCs out. So the we couldn't go public. The VCs took a minority stake in the company, and then eventually they wanted a liquidity event, and they got all nasty with me.
And so anyway, we worked out an arrangement because I'm a cashflow guy, right? So I'm always a cashflow guy. So we came up with an arrangement, and we came up with evaluation, and then I bought them out. Okay, and then I eventually bought all my investors, including my friends and family, that actually got a great return, because they waited with me for me.
But the moment the VCs stepped away, I had a bit of a relief, but I had this yearning to help other entrepreneurs in Portland. And I started mentoring one group, one of the guys was in my yoga class, and there were a bunch of Intel engineers that had some idea, but a little necklace device to wear was a fall detection device that somebody could wear. And you know, that whole ad about here, I fall and they came, get out. They wanted to solve that problem.
Right? So they're working on this, I'm mentoring them. And I got approached at the same time by a bunch of people who wanted to, we're starting a chapter of tie. I don't know if you've heard of tie, it's the Indus entrepreneurs.
It was started in the mid 90s, by very successful, including Vinot Coastlaw and others that were early Indian entrepreneurs that come into the United States in the 60s and the 70s. And they wanted to give back. And some of them, you know, took their company's public. And part and actually Hotmail actually came out of that mentoring in the early 90s.
And the whole idea behind tie was to get entrepreneurs VC ready. So mentor them, coach them, and they get them ready for investment. So that chapter can't come in, it became very successful global organization. They came to Oregon in 2007.
I joined as chair of mentoring and the markets crashed. And the number one question I would get was advice is cheap and then show me the money. There's no VC money left around here right now. So that became a catalyst.
And I'm actually a female entrepreneur who was a mom in the same school as my kids were. She was an ex-into employee and started this company and suckered me in as a mentor first and then asked for money. And that led to me starting an angel group within tie called tie organ angels. And we started making angel investments in 2008, nine, 10, 11, so on and so forth.
And the differentiator for us was because we were mentors first and angels as kind of like a secondary thing, it became what we called mentor capital. So I coined this term called mentor capital. So that's what we're doing. And we built some very helpful, some very successful companies in Portland.
Most of them sold and got great returns. So that work got noticed by a local endowment called the Meyer Memorial Trust in Portland. And the chief investment officer was familiar with ties that come out. She approached me to see if I want to start a fund.
And that led to the creation of elevate and with a very specific thesis of investing in overlooked and underestimated entrepreneurs. So they weren't people like me because I had a hard time raising money. So what we wanted to do was go into these ideas, early stage ideas where founders had difficulty finding investors or networks and then help them with both. So fun one we launched in 2016 became very successful.
We had some big hits, very early big hits. We did some, we did, we were a diverse. We did some medical device, some health care tech. And so that's really how we got started in 2016.
Then we raised a second fund during the pandemic. So we started from 10 million to 40 million. Now I'm raising my four third fund. And we just closed one investment out of that.
It's in the I company in Cleveland. But our thematic approach of investing is look in verticals. So healthcare is one of the verticals that we look at because that's my background. And we're only doing AI and software deals.
We don't do medical devices anymore. That's a long tail. We're not experts in able. And what we provide is not just not just the money, but we have like a big network of LPs that are exited entrepreneurs.
Some of them are actually medical professionals as well. So medical professionals who are entrepreneurs. So what we provide is not just the funding, but also the networks that they might need for either building syndicates for investments, also hiring people and getting customers. As we sort of wrap up here, what's your one piece of advice that you would give to someone looking to raise money or someone with a product or an idea, even an idea of a product?
Make sure you have really good customer validation. Because that to me is the key. If you walk in, that's your skin in the game. That's the number one advice we give.
We actually teach this to high school students. Customer validation alongside some design thinking like this, you know, flushing out the idea. But that market analysis and talking to customers and then feeding that to a potential investor. That data is worth gold.
Because that's what people get. Obviously, you want to assess the market, you know, like a tiny market. So those things are important. But the most important thing is, you know, how validated this is this idea.
And subject matter experts tend to be more fortunate in raising money for that reason, because you come from industry. So if you've already experienced the problem yourself and you validated with 10 other people, that would be my advice. Well, then, thank you so much for your, again, for your generosity today and giving us this sort of beautiful narrative arc from coming here with $1,000 and your father having lettuce and his his his shoes and you develop a company to then getting back through your own VC. So this has been quite a journey.
And I think I love the I care industry. I love working with our doctors. I never became an ophthalmologist, but, you know, I get so much joy and pleasure just helping helping, you know, the the I care community with whatever whatever we can provide. So it's been and you've helped a ton.
And so certainly, thank you again for taking the time to be on our podcast. Nitten, you're you as someone wants to contact you, how can they how can they contact you? They can give my cell phone, they can text me 503-799-5200 or they can send me an email at nitten at elevate.vc. That's my elevate email, e-l-e-b-a-t dot vc.
But yeah, I'm game. I'd rather get calls from people that are your listeners versus all this garbage. I was going to strike that for the to be edited out, but maybe we'll keep it. You could reach us at ai and icare.com.
This is Rahan Ahmed again for the AI. I care innovators podcast. Thank you everyone for listening and we talk to you next time. Alright, thank you.
Thanks.