The Rise of AI Agents in Healthcare | Omri Yoffe, Vi Labs

We have one of the largest world data sets for clinical, behavioral, and operational data. 190 million lives and 1.2 pabytes computer data.

That's insane.

Today's guest is Omri Yoffe, an entrepreneur building V, an AI execution layer, revolutionizing health care and life sciences through artificial intelligence.

One thing I noticed in the healthcare industry is too many parties involved. The insurance, the government, patient, the doctor. Healthcare is still, I would say, failing. the people it's supposed to protect and I think AI is the best cure.

So you're saying in a decade from now we'll have robots doing surgery. What was your early interaction with AI? Because AI became mainstream in 2022.

We wanted to create an AI companion that will live within your phone. We did that. We delivered the product and then got a very important lesson.

Vlabs currently valued at $1.64 billion. How does it feel to be a unicorn status? I think

today on the podcast we have Omri Yoffe. So happy to see you Omry.

Same here. Thanks for having us.

So I wanted to start the podcast on a very light note. So what's wrong with the American healthcare industry?

Healthcare is still I would say failing the people it's supposed to protect sometimes unintentionally. Um people are diagnosed but don't getting the right treatment uh with the the best next next action fast enough. drugs are very slow to market given some complexity. We can go into it later in terms of clinical trial completion. Um and you know and and grandmothers don't have access to affordable healthcare although they can but it's very hard to connect the dots for people to get access to the system. So you're talking about a highly complex sometimes misaligned incentive system. You've had a very multiaceted career going from an air force pilot to you you started a biometrics like wearables company and now you're in enterprise like healthcare. So can you talk a little bit about the transition? I know the healthcare is like the common theme here.

Sure. So I got into the health mission in a very uh abnormal uh random way. Uh as you've mentioned I come originally from the aerospace domain and founded my first business as an R&D powerhouse called clerka to basically deliver different types of avionics command and control systems heads up display uh to the largest aerospace organizations out there very high stake uh environments if you will and the idea I'm a bootstrap guy so a big believer in trying to build things in a self-sustained way

and in that case started the the early years of clerk had generated positive cash flow and looking for more impactful scalable solution using technology. Um, one of them was called LifeBam. We lost one of our best F-16 pilots in the Air Force, got a GLO. He basically pulled high G's, passed out and crashed. And we've built an application, a solution uh that was uh designed and still does saving lives of pilots, special forces, and astronauts during unique situations, fatigue, hypoxy, hypoxia, um um passing out and different types of physiological weaknesses. We built and solved it uh uh by the end of 216 and I got in love into the problem of precision health and predictive health. I never looked back. Founded V did uh a material pivot in between. We can talk about it later. Same mission but went from a large scale direct to consumer to enterprise and uh very obsessed and very passionate about the health mission.

You were pretty early on so many things, right? Like we just saw Whoop got valued at 10 billion dollars. There's Aura Ring. There's uh Eight Sleep. There's so many interesting companies in the wearable space. How do you think about that industry? And because you were so deep into it.

Sure. You know, it's crazy. First of all, you're right. Timing, I don't want to say it's everything, but um I think what I learned and the the a good uh lesson in humility, if you will, uh is the following. We from day one, we wanted to create an AI companion. I'm talking the early early days of V that will live within your earphones, within your wearables as you've mentioned, and within your phone and uh guide you to a a healthier life journey. We did that. Uh we got to $56 million in sales. You can check it out over the web. We had the the fastest and largest Kickstarter campaign out there. We deliver the product a full product market fit and then got a a very important lesson uh from the Apples and Googles of the world when they went in with our health kit, Google feed and so forth and cost of acquisition of our solution went to the roof. Um, and I had two options. Either knock our hand in the wall and uh and you know and try and uh fight a very non uh unwise uh fight or saying we want to make people healthy. Let's use our edge and advantages and learnings that we got out of that uh wearable solution uh into the enterprise world. Pivoted very aggressively. So I had 53 people. I needed to release 50. M

um some of them came back by the way uh adapted pretty fast into the B2B world and uh never looked back.

I was watching on YouTube your one of your uh presentations and I I one thing I really um took away from that you're you're deeply technical person. Uh can you talk like about that a little bit because you're also you also have an MBA but like

I'm a very obsessed uh deeper not wider guy and if there's something I want to be fluent with I do invest the time and surround myself with the right teachers and real life environments to be fluent in it. So I don't have an engineering degree but you are right anywhere from systems design, hardware, software, data science, actual product design. I am uh above the average uh in terms of being in a room and be able to have an educated discussion about it and and pressure test some questions versus let's hire a bunch of people and and let them fly. Um and [snorts] I think in today's world it's a must if you want to have uh the basic uh intuition into what's feasible and what's not. Uh you must be fluent with those data pieces. We will talk AI native in a second. There are some new leverage coming in. Uh but I think traditional self-arning will not change.

Uh what was your early interaction with AI? Because AI became mainstream in 2022 and you were doing this in 2016.

Correct. So we had back in 216 217 machine and deep learning started to be a thing from the MITs of the world to the Techneon and some other people that worked with me on other platforms. And for me it was clear that I want to try and prove a thesis on can we mimic human practices and human mimics into an AI companion. Now it sounds cool. Now you didn't have Siri. You didn't have you had like basic I don't know if you you remember a company called nuance based in Boston and others. So we had basic TTS text to speech and automatic speech recognition and we had machine learning models that can predict human activity based on tagged. You need to tag the the data in order to do so. So in order to build a health coach we needed to record and Vix is uh behind the scenes here uh helped us doing it. We basically worked on a human voice. We wanted the voice to be fully uh uh human and authentic versus synthetic. So we recorded tens of thousands of uh pieces of of voice snips. We you know interviewed many voice actress. One of them lives here. Her name is Crystal. Correct. And we we built a coach based on her voice. And it was the only coach in the world that had predictive analytics and the ability to trigger the right commands and conversational parts but also have stitched in, you know, conversational sentences, numbers, names, and so forth. So, I'm giving you a a non-scalable solution for an AI problem. Uh, today it's a it's a it's a no-brainer that you can basically use Grock or OpenAI uh to to speak with our TTS voices. Second example, as you mentioned earlier, we didn't have NLMs. So every model they need to deploy, you need to take uh what we call deterministic model, specific data, tag it, train the model, see the outcomes. In today's world, you have aic systems that can do all of it by themselves. So there's a massive jump people I think underestimate. Uh but I do think this traditional knowledge helps to to determine what's right and what's wrong. So, here are some exciting numbers uh from today, present day 2026. Uh Vlabs has raised over $160 million and was currently valued at $1.64 billion. How does it feel to be uh a unicorn status? To be completely direct, it doesn't feel uh we just had a town hall meeting with my team um and I ended it with a very simple slide called stay focused, stay humble. Uh, and it's not an empty statement. I think if we would have done what we're doing for the sake of valuations or money, we'll probably sell the company by now or go to the beach with some great restaurant, hotels, and and movies, which I'm not against. I think it is important for people to to enjoy their life. But in V's case, I genuinely believe that it's day one always. The type of people we hire and train and grow are people that are here to improve their own version and do exceptional things regardless if you have another zero to your value or not. I think practically speaking valuation helps to do two main things. You want to retain the world's top talent. So you can use this as a way to get every single employee as a shareholder in the company and to think and drive enterprise value. Number one. Number two is shareholders. you you need some measurement and returns for people uh to to make sure that they can recognize success in their minds. But for me personally, my wife is joking that uh she never knows if we are up or down because I I type of, you know, push the same level. Um and and I think it's very important for the rest of the team as well.

Got it. And uh what's the elevator pitch of Vlabs? Uh I know some people have called you like the palunteer for healthcare industry. Uh there's a lot of um can you explain in in your own words?

Happy to. So in short V is the AI execution layer.

Okay.

Basically helping healthcare and life science enterprises to turn their data into next best actions for patients and for care teams.

These are hospitals.

So exactly. So three quick examples. We help the most important clinical trials out there to accelerate and complete faster. To your question about hospitals, we help patients and physicians from a surgery to an appointment uh to basically take their next best action, second opinions to move their health journey in a much more precise, predictive and fast way. And we help optimize supply chains and operations. So staffing, scheduling, you know, uh different types of resources management. And the way you should look at us is we are you used as an example. We bring three main things to the table that differentiates us from the Palanteers and and other enterprise AI players. The first is our data edge. We have arguably the largest of one of the largest world data sets for clinical behavioral and operational data as you've mentioned 190 million lives and 1.2 pabytes computed dates. It's a massive data set. That's for context. I actually googled it. That's 1.2 million GB of data.

Yes, sir.

That's insane. It is.

How did you go about collecting that data?

As you mentioned, it's not an overnight success. You need to go day by day, year by year, and gather and and and make this data, by the way, fully privacy safe. So, we anonymize all of the patients, but we gather lots of learnings and knowledge about the health space. Number one. Number two is we deploy and productize agents and models specifically specializing in the health domain. So we we call it the vertical AI application there. And number three, we have skin in the game. Our business model is very straightforward. We win only if our customers and their patients are winning and we have a way to price and keep ourselves accountable for their outcomes.

So it's not a traditional SAS model like uh how would you explain your business model? So you're spot on and it's it's it's actually a very critical point. We for the last five plus years are selling value and this is not because of the current hype of selling ROI. Uh we always believed that selling on a seatbased level is the wrong incentives. Uh you want to be completely aligned between you, your customers and anyone that is part of the valuation including obviously the patient itself. So in super short I'll keep it at 30 seconds. We go into a customer, we underwrite and map out where we can help. We can reduce the cost of patient activation or clinical trial acceleration. We can increase the amount of usage in a healthy service for a hospital or a provider or we can save you tens of millions of dollars in operations. We bake it into our pricing model. We go into a pilot of three months with a control group. So you have fully transparent with V and without V uh u control group versus action group. We prove value and only when we prove value we turn it into a flat fee monthly fee based on this ROI and there's full transparency for the client. So if we don't deliver after a year after 6 months we can stop and reunderite to make sure they do get the value they need.

That's a much better business model right now because SAS is going through a bit of an apocalypse. Uh have you thought about being partners in these businesses and taking equity?

We did. We definitely did. You want to make sure you're staying the Switzerland of the industry if that makes sense. So you know my job in life is to be as impactful and scalable. If we want to create what we call health abundance in the next five, seven, 10 years, you can do it if you have a an over incentivized uh across a specific pharma company or or healthcare company. So we make sure to get upside but we never took equity. But you're asking the right questions. Many companies are saying, "Guys, we feel that you're an extension of our technology. You are an amplifier. We want you to think enterprise value. We usually turn it into cash normalized by MR because you also don't want to get a $10 million check, a one-off. You want to create predictability uh should the company at any point will want to go public which is definitely not the case right now or just want to create predictability for investors or shareholders you want to normalize it if you will over two to three years ahead

and uh you're I'm guessing you're not competing with the open AIS and the clouds of

we work with them those are our one of our partners so vets on top of existing data stack so you keep your data bricks you keep your epic you keep your uh viva Um and we also leverage our own proprietary models but third party models like as you've mentioned anywhere from uh the uh the open AIs of the world and entropics to the more local models and private models and and uh open source models. We use those models and productize and train them based on our data and expertise. So they like it. They are think about them as the OS store. We build the actual application for the end client. I think the good thing about OpenAI they're they're infinitely scalable. How scalable is your company like and how many countries are you targeting?

So again we as mentioned we need to stay focused on humble. So I don't want to throw uh empty statements and numbers out there. I can give you uh nonV related numbers. So first of all only the healthcare and life science market by itself is [snorts] 19% of uh US GDP and 12% of global GDP. So we're talking about trillions of dollars compounded. Number one. Number two, the fact that V is a Switzerland. I don't focus on a specific clinical or population expertise. And I also cover the full value chain. V activate new patients. Ven engage existing patients. V operate for operations. I don't want to say endless time, but we have years of growth ahead of us. And every time I'm being asked by board members, Omry, what you were doing now, we might be able to do it for, you know, for mobility and governance and defense. I'm here because of the purpose itself, because of the mission. And I think we're fortunate enough to have enough big of a temp to focus on it for the coming years.

Uh I saw in an interview you said you use 20 LLMs. Can you talk a little bit about that?

Yeah. So I don't want to treat it as like a hard-coded number, but what you need I think to to here on this one is the following. We are cherrypicking um deterministic models and probabilistic models which are the the LLMs of the world. Usually it's 15 to 20 chosen models at time and the reason is you want to pick novelty models and specialize and productize them versus being a supermarket and a reseller of models. And it's very nuanced. It's not like that every quarter we change all models. It's a it's an incremental okay we now there's a new LLA model that coming out uh like the OPUS 4.7 is better for internal coding and agents great uh maybe there's a different mistral model that we can deploy for operations let's check it but it's a cherry picking piece versus a hardcoded list

yeah I still uh I've been using a lot of agents in my businesses and you still need a human in the loop

for sure

how do you think about agents

first of all I I think you're spot on and I think we you on the two ends of the value chain. On the V side, we have QI people. It sounds obvious, but we believe we have a very flat AI native approach. I can explain how we do it if if you're interested, but we basically have an IC culture from day one and AI native is amplifying it. So, it's not that we have 10 humans per employee. We have small squads, very prominent and fluent AI people that usually have traditional software capabilities, but you also have QA testers to your point. There's lots of hallucination and probabilistic potential mistakes. You need someone to work off the workflows. This is on the V side. On the patient side, you have moral judgment and compliance problems.

So, you need agents to do the heavy lifting. I'll give an example. um you went to a United Healthcare coverage or to an HCA hospital, um you were diagnosed, god forbid, for a kidneys problem, even if there's an agent that once you went out of the doctor, triggered the next best action and gathered and scheduled the next second opinion appointment and got the payment from your healthcare provider and employer and sent you an update about your journey and what's expected. Great. You still need a human to intervene. It can be a nurse. It can be an engagement specialist. First of all, to show empathy and and for the person to understand that it's being managed by humans with the right values and care, but second, sometimes the AI will do stupid stuff. I can give examples. So, we're moving into a world that you have massive heavy lifting by agents, but also human authenticity, judgment, and compliance by humans.

Uh, one thing I noticed in the healthcare industry is too many parties involved.

True. the insurance, the government, the patient, the doctor,

the healthcare provider. Uh, how do you think about that? It's like this is like for someone who studies game theory, this is

7D chess.

I don't have a a silver bullet answer here. This is a an orchestrated political uh regulated um capital sensitive and and commercial battle. But I think there are two things that if they will happen, we have a massive opportunity to solve what you're saying. I think the best future leaders if they can run on the ticket of saying we're going to create one source of truth and a connecting grid that Manavan Omry will be able to sign a consent and all those parties that you've mentioned will see our main ID member clinical claims and health journeys from dentist to god forbid lukemia. Massive opportunity. So that's number one, doable but requires leadership uh and execution. Second is you need CEOs, CTO's, COOs to turn their organizations to be AI ready. Um you know it's the data bricks, the snowflakes of the world, the clean rooms of the AWS's and Google clouds. This is a must. Uh and and I think once those two things happen, health abundance could be achieved. lots of work to do.

Uh kudos to you by the way for building in such a hard sector. I think a lot of people just avoid healthcare because of the bureaucracy in this industry. Uh I want to touch a little bit about on the healthcare abundance and let's make some crazy predictions. You know we'll come back to it 5 years later maybe laugh. Um yeah tell me some predictions because uh things are changing so much like with full self-driving car like we have so much abundance right now. Yeah. So tell me tell us about abundance in healthcare.

So I think there is a way and I don't think it's crazy. I think we need to be very concrete and ambitious about how to measure health abundance. And I'll use three main metrics. The first you want to get to a precision level that you don't do, sorry for my French, but stupid mistakes that kids, parents, grandparents are dying or suffering for the rest of their lives just because there's lack of precision. And I don't know if it's 100%, but it's not 80. That's for sure. We want to get to the 80 90% precision level. So your health care system is able to drive the next best action to the right second opinion to the right appointment to the right surgery and so forth.

Second is you want to be predictive. Time is critical in health. So again, I'm not talking 100, but I imagine a world that Manavan Omry a minute after they went out of the doctor at least at the same day are getting a very clear GPS journey that is already predicting what are at least the avenues and junctions that they need to do. And I think it's very doable, very doable. I can give examples. And lastly is affordability. It doesn't make sense that most people in the world, not only in the US, don't have access to affordable care. And I think AI is the best best cure and the best ingredient because it lowers the cost of opportunity. It lowers the heavy lifting of actual humans needed to do repetitive tasks in a very cost non-cost effective way. So I think you're going to see better decisions time uh being shortened and most important I'm I'm very confident given where AI is heading that over the next five seven years uh most citizens in the western world in the advanced world using AI healthcare is not going to be the main burden for them as it is today. we'll have robots doing surgeries

all day long.

Yeah.

All day long. and you'll have abundance of access to data. So you don't need to think about it that way. uh if I have god forbid a very nuanced clinical problem, I don't need the best Sloan or Mayo Clinic doctor, it's fully democratized and each one of us across this room will have the exact same access to those decisions number one and the exact same access to the care at least executing or facilitating your career navigation which is people don't understand that's the key piece. It's not the drug or the science problem. The science is there. It's how do you orchestrate and navigate the patient to take the right decisions.

Got it. Uh and I kind of want to end on the note of what's next for V like what maybe what roles you're hiring for or what's the next big vision like for

so we want to build what we believe is going to be the most important and generational healthcare and life science enterprise solution in the coming years. And I think the two areas that I'm most excited about, the first is I'll call it context as data as a catalog. We're building a very powerful AI native infrastructure for our teams to execute in a disproportional way. Meaning I want to serve, we serve today over 100 of the largest enterprises out there. I want to serve 300 with 250 people.

So there's lots of investment being made behind the scenes. We can do it maybe for another chat. But this is one area. How do you build let's call it a team operating system. So a data scientist, a squad leader, an applied AI engineer can be disproportionately effective for a very large enterprise. Number one. Number two, human touch is critical in AI. So you're going to see vhiring. And I actually want to uh use this uh uh opportunity to share it w with u people that are watching us or listening. We look for extremely competent for deployed engineers. You still need people if you work with a $50 billion healthcare or pharmaceutical organization, you want people to really live through their strategy and their KPIs. It's not 20 people. It's one or two people per logo. But we're looking for people that has good technical and operational understanding of the customer, but also very fluent on AI and the technical stack. Those are the two areas I think are going to unleash the next chapter at V.

If you're a smart engineer and want to be part of a unicorn like you know [laughter] go to v.co. So where where can people find you?

So as you've mentioned we are at v.co. uh we're publishing most of those roles out there. I'm available on LinkedIn. So, whoever wants to reach out, I'm very responsive to any question, anything that we can be helpful or valuable. Um and you can follow us on LinkedIn, YouTube and X.

And one thing I want to ask you, what's one advice for any founder, entrepreneur watching and want to be the next Omry?

Never give up. Don't don't you know don't be afraid of shame, humility, optics, uh focus. Make sure that you as Einstein said, you know, uh intelligence is the ability to change. So make sure you change over time, but never give up and uh over a compound effort and time uh you will uh you will prevail and will win.

Thank you so much. Thank you for coming on the show, Amry. I really I know how busy you are. So thank you

for sure. Thanks for having us.