Doctronic CEO: 20M AI medical consultations, prescription renewals in Utah, and the race to be the AI doctor

Let me tell you about MongoDB. Choose a database built for flexibility and scale with best-in-class embedding models and rerankers. MongoDB has what you need to build what's next. Our next guest is Matt from Doctronic. He's the co-founder and CEO.

Here we go.

And we're going to welcome him to the stream.

What's happening, Matt? How are you?

Can you hear me? Not yet.

I don't know something happened with the AirPods.

There we go. There we go. AirPods. AirPods are rough. They create this split-second delay. Now it feels like

Oh, I didn't know that.

We're face to face now. Uh, how you doing? Great to meet.

Yeah, great to meet you, too. Thanks for having me on.

Uh, first time on the show. Would love a quick introduction on yourself and the company.

Yeah. Yeah, sure. So, I'm Matt. Um, I've been building startups for a long time. 25 years at this point. Built a big

success.

Yeah. Right. Big built a big e-commerce unicorn called Moto Opera Andy a long time ago. Y

a lot of startups computerized from Carnegie Melon. Um, doctronic is the number one AI doctor, right? So got 20 million patients or or 20 million consults so far, 2 million patients.

Um, wow.

And we're an AI doctor where anybody can come for free, talk to AI, etc. And we have a tele health practice behind it too. So you can talk to AI, it's going to diagnose you at the end of that. You can take all that information, full diagnosis, full treatment plan, etc. Uh, and talk to our doctors. So it's a a pretty straightforward simple process. Um our doctors are are available 247 licensed in all 50 states.

Uh when when did you when did you start the company and and what is uh what's been the primary kind of distribution method? Is this something that people just they end up they have a pressing need. They want medical advice, guidance, they're going to find you. Uh or are you selling through you know how are you actually getting 20 million people using the product?

Yeah. So we're direct to consumer. We launched in September of 23, so been around for a while. Actually, when we launched, there was this open question, uh, you know, did anyone want to talk to an AI about their health? It was kind of an unknown. It was, you know, early early in the the GPT4 days. Um,

yeah,

and it turns out it's a resounding yes, right? People really want to talk about their health. The other question was, could you make an AI accurate enough to practice medicine, right? Um, and it turns out you can. Uh, you can't just use the foundational models. You have to do a lot of a lot of uh work with them. Um, and so we're direct to consumer. We get a huge amount of our traffic from just organic search. Um, you know, we get 100 150,000 people hitting the site a week. Um, and then of course we're doing social networking, paid stuff, etc.

How much more do people when when when people [clears throat] have a doctor in their pocket, do they are they going are they effectively talking to a doctor or an agent that's acting as a doctor? Is it is it 50 times more? Is it a hundred times more than they usually would? Because I mean I think everybody's experience like you know researching it was you know doing a search adding Reddit to the end. Now people might try a foundation model and it's obviously giving a disclaimer. Well we're a foundation model. We can't actually give you know medical advice.

Do you have to put a disclaimer like that or or still or

how does it actually work?

Yeah it's super interesting. So so first off let's let's talk about like how often people use it. So, I think the the stat is the average American sees their primary care doctor three times a year, which is, you know, kind of crazy. And of course, there just aren't enough primary care doctor

for the guys out there. They're like three times a three times in my life.

Yeah. [laughter] Right. I haven't seen mine in years. My wife is a doctor, which makes it a little bit easier. But yeah, you know, um but but my wife doesn't want to help me anyway, right? That's that's not something you ask your wife. [laughter] But but yeah, so uh people are using the system multiple times a month, right? So, we're close to weekly active users for a lot a lot of people. Interesting. So, the question as to like what what would happen if you had a real doctor in your pocket is I think you talk to them about everything. You talk to them about how you wake up in the morning, your back is sore, you know, you have herniated discs. Is it worse than a couple days ago? You don't know, but you talk, right? And it slowly gathers all this information about you, right? Um, so so it's certainly changing the way people are are interacting. uh as far as disclaimers go and such, we are so our AI for most situations outside of the state of Utah right now doesn't practice medicine. So it's going to talk to you and it's going to always say look

in Utah in Utah you guys it does practice medicine.

Wait, what?

Yeah, that's the cool part. [laughter] That's what we announced today. I'll tell you about that in a second.

Yeah. Yeah. Yeah. So just so outside of Utah

Yeah.

You know, let's say you're on a diagnostic pathway and you go through the diagnostic agents, etc. Right. The AI is going to kind of come to this consensus and they're all going to say, "Look, it's probably one of these four things, right? And here are four treatment plans for each of those and here's a note that you can take to your doctor and we're not practicing medicine. You can also take this to our doctors. This is all just for you to take to your doctors. It's a tool to help you communicate with your doctors and it's more accurate than chatbt and Gemini and all those because we're feeding it all sorts of you know this massive corpus of medical documents that our doctors have written for the LMS and we've got lots of agents. We can get into details some other time, right?

Um but it doesn't practice medicine

at all. It's going to just do you think that uh this aspect of Gen AI is still kind of underrated by the world? I had a kind of a weird uh cool moment this week where uh my my one-year-old wasn't sleeping that well. My wife was about to hire a sleep consultant that we had hired for our first kid. She started talking with an LLM and it basically within 24 hours had like got the one-year-old on a sleep schedule that was working again. And that was such a cool moment because I was like, we were happy. Historically, it was so worth it to hire a sleep consultant because if your kids's not sleeping great and they can and you get them sleeping great, it's like life-changing.

Um, but now but most families are not going to spend $1,000 to have some expert come and coach them on sleep schedules and things like that. So, it feels like underrated at a society level. People just see the slop in their different social feeds and they're like, "This is dumb. Why are we building data centers?" Um, and yet

I think you're right. I think it's totally underrated. I read the other day that 20% of Chachi PT traffic is health related and for for things like that helping you sleep train your child.

Yeah.

Awesome. Like use Chat GPT. It's great, right? But when you actually want to start getting into medicine like like real actual medicine, Chhat GPT can't do that. It can't. I mean, even if it could, right? It probably shouldn't, right? But it's not going to help you understand a prescription and what the dosage should be and what the schedule should be. If you have high cholesterol, it's going to say you probably need to take a statin and then you need to get some exercise, which is great and true, right? And that's just helping you with your health. That's not practicing medicine, right?

Yeah.

Um and which, you know, again, outside of the state of Utah, we're not doing that either,

but we we have a partnership with the state of Utah now uh with um with the Utah AI Learning Lab.

And so that launched a couple weeks ago, but we just announced it. So, our AI is allowed to legally renew prescriptions for residents of the state of Utah with no oversight from doctors.

Wow.

I see uh uh Tusk Ventures on the cap table and I think the goat

I mean Fe Lee amazing. Um but uh I Fe Lee signals like you know what you're doing technically. Uh Tusk Ventures signals like this is going to be a city by city, statebystate like battle. At least if for those who don't know, Bradley Tusk uh worked with Uber and and did a lot of the on the ground athlete hall of fame.

Exactly. Regulatory athlete. And so I'm wondering like do you think that there needs to be regulatory change on a state-by-state basis? Is there a federal preeemption discussion in the mix? Like where does all this go? Um and and what what are there any like keys to the success for you uh on the regulatory side?

That's an awesome question. So historically, not even historically, currently, states regulate the practice of medicine, right? If you want to practice medicine in a state, you get licensed by that state's medical board.

Yeah.

We believe that what we're doing is practicing medicine. Obviously, the state of Utah agrees. Yeah.

Uh and so we have been

kind of mitigated from the laws in this AI sandbox with Utah that prevent non-licensed doctors from practicing medicine.

We think this applies to other states, too. And other states have regulatory sandboxes for AI, and we're talking with them as well. Uh, it's just that Utah is super innovation friendly and forward thinking. They're the first. They're the best. They're really good at this. The state's right.

Okay.

What What do What do doctors think about Doctrronic? Are they happy that they're patients? Because the patients coming in, they're informed that the Doctrronic already kind of summarized the symptoms and the conversations, gets them up to speed quickly, helps them process more, see more patients themselves. We we see like a 10x efficiency in our doctors in treating patients because the AI has done all of the work gathering things. My co-founder who is a doctor, I'm not, but he is. He says it's like having the best chief resident in the world seeing every single patient for you. It just it just does all that work, but then it says, look, it's it's one of these four things. You you figure it out, but it's almost always the first one, right? The AI is just that good.

Yeah. Yeah. I I've had a couple funny experiences lately. I went to the pharmacy, picked up some uh medicine. My wife asked me to ask the pharmacist how to use it. And I watched the pharmacist, you know, just use an LLM to get the answer. I was like, "Wait a minute. I I have a more expensive plan. I should just be doing this myself." So, the demand is is

Wait, you're not even using premium plus?

Yeah. Wait, wait, you didn't even run the deep research report on this one. Um, but do you think that there's is there going to be some sort of stop gap where there's some sort of hybrid where you have a human in the loop to accelerate progress? Do you already have that? What what role do you see as like uh building out the the the hybrid uh portion of the business in certain states that are maybe less uh aggressive about moving quickly on this?

Yeah, there's certainly a place to use those efficiencies you get from AI in the states that are are not able to move that quickly. Yeah. Um, our hope is that that hybrid approach won't last for long because we will prove that this is effective and works really well. It saves the taxpayer money. It makes people healthier. It reduces costs, right? The healthare system is enormously complex and far too expensive. There aren't enough primary care doctors. No primary care doctor wants to renew prescriptions. Nobody went to med school saying, "Gosh, I hope I can renew prescriptions." So, I I I sure hope that we'll be able to move faster, but there are definitely hybrid approaches. We're already doing that with a bunch of our our our video visits in in other states that don't allow for this.

Sure. Sure.

Uh you have a background building in e-commerce, so a little bit of a tangent, but I was curious what you expect to see at the intersection of of LLMs and commerce. This year specifically, it feels like last year people started discovering and doing a lot of product research using LLM. It feels like this year will be the year of like

right

what's that?

Like I said, "I sure hope we're shopping in chat GPT." Yeah.

Or or or Gemini, right? I should be able to buy something from there.

I shouldn't have to go to Amazon. I should just be able to click buy.

Yeah.

Yeah.

Love to see that.

Yeah.

Yeah. Something as simple as that. Amazon's like, "Actually, we like our search ad revenue business quite a lot.

They'll figure it out. [laughter] I'm sure they'll have their own LLMs or something like that." Um what what do you think about uh like the actual feedback loop on the I feel like there's like the value of medicine is not just treating the patient but then filtering those results back into future research efforts to update on what's working not both on the individual patient level and also on the societal level. Um how do you think about improving the system over time? Um it's a little bit different than just just coding in my opinion. But uh how are you thinking about uh you know successive interactions with with the with the patient?

Yeah. I mean you know AIs can learn more and remember more about patients anyway. And the funny thing is you know we measure NPS with the AI and with our human doctors

and we have great human doctors right like we we we pay a lot. We hire the best.

Yeah.

Everybody still likes the AI a little bit more. Mhm.

Uh that said,

I don't think that AI is going to replace that deep human interaction.

Yeah, of course. [laughter]

Right. Like like there's there's this easy 80% of these wrote interactions that you have with a doctor where you're just like, "Look, my kid's sick. I need an antibiotic." Yep.

Let's just figure out which one is most appropriate. Let's get the AI to handle that.

Uh I I don't you know, specialists diagnosing chronic conditions.

Yeah. Yeah, people are always going to be

at the same time with that if if the kid's sick, they need antibiotics. I I feel like you should still maybe bring them into a clinic, get some some data that can't be accessed on the phone. And so, doesn't that lead you to some sort of uh like B2B scenario or somewhere where you're acting more as a a co-pilot than uh you know, a BTOC company? Do you think you'll get there or is that something that's like adjacent?

I I think that's adjacent to where we are right now. I I guess I guess the way I think of it is um we are pretty well positioned to handle anything that can happen during tellahalth visits right now.

That's kind of our goal, right? There there's already this a bit of a line drawn that said these are these are things that can happen with tellahalth.

Uh and so I mean you're right there are certainly cases where you just need to see someone. We we had one the other day. I mean they happen all the time but we talked about one today. Someone had an appendicitis, right?

The AI said, "Look, you you need to go to an ER. You have an appendicitis and it's about to rupture." And the person still chose to see our doctor. And it was great because the doctor said, "I totally agree with the AI, but this is pretty emergent. Like, you probably should go to the ER and they did." And they were taken care of and and it happens. Yeah.

Um, so we're not going to try to uh to do those things that we can't do via a tele visit anyway.

Yeah, that makes sense. Uh, well, congratulations on all the progress. Thank you so much for coming on the show today.

Yeah, very cool.

And I hope you have a great

Are you the Are you the uh first AI licensed to practice law or

or sorry, not law, medicine, medicine.

Yeah. Yeah. Yeah.

Wow.

That is that is wild. That uh

uh I feel like Yeah, I I imagine in 5 years we'll look back on today as a as a very big moment. So, congrats to you and the whole team. It's great to meet you.

Congratulations.

Thanks so much.

We'll talk to you soon. Goodbye.

Appreciate it.

Let me tell you about Gusto. The unified platform for payroll, benefits, and HR built to evolve with modern small and medium-sized businesses.

Gusto,

let's send Lulu.

We'll bring them in in a couple minutes. In the meantime, I'm going to tell you about Cognition, the software. Uh the uh they're makers of Devon, the AI software engineer. Uh crush your backlog with your personal AI engineering team.

We have uh some breaking news.

What's the breaking news? from two hours ago from XAI. They've raised a $20 billion series E

20 billion.

XAI completed its upside series E funding round exceeding the 15 billion targeted round size and raised 20 billion. Uh investors include Valor Equity Partners, Stepstone Group, Fidelity Management,

Antonio Gracio, Investment Authority, MGX, and Baron Capital Group. What is Baron Capital Group? Is that is that Baron Trump? No.

Since 1982. Okay. [laughter] Not Baron Trump. Strategic investors in the round include Nvidia and Cisco. Let's give it up for Cisco. Investments getting in the mix.

We're still working out the kinks with this gong. We're figuring out that that was that was a clean hit. That was a clean hit. Uh but we're figuring out the new kind of method how to strike it effectively. Uh but that was a great hit. Uh 20 uh XAI says 2025 was a year of breakthrough momentum where the XAI team advanced a multitude of key initiatives including data centers, Grock 4, Grock voice, user metrics, and Grock imagine. I'm very excited to see what Grock um pulls off uh Grock and the XAI team pull off this year. They are highly highly motivated, highly capitalized, crushing it on the infrastructure side and a lot more to come. Well, before we bring in our guests, let me tell you about Phantom Cash. Fund your wallet without exchanges or middlemen and spend with the Phantom Card. You can see Bitcoin is back up $93,000.

There we go.

Uh, and let me also tell you about Railway. Railway simplifies software development. Web apps, servers, and databases run in one place with scaling, monitoring, and security built in. And with that, we are ready to bring in our guest. Um Elon is saying that we've entered the singularity. So he is fully AGI pill, fully singularity pill,

Lulu singular,

GDP grow% at at with GDP growing 100% a year. 20 billion for XAI in series E seems like a steal.

How about 200 billion?

How about 200 billion? I I'm dying to see the the user metrics. Eventually people will pull out um we'll pull out Why are we doing bubble talk?