Subscribe
Share
Share
Embed
Vic and Marcus talk with Carlos A. Rodriguez, MBA, CEO and Co-Founder of Humanate Inc., about his career from Bain and health system leadership to building a $40M SaaS startup and now launching Humanate, a company focused on agentic AI for healthcare. The discussion covers the origin of “Cassie,” a multilingual AI clinical concierge with ASL support, how staffing shortages and labor inflation shaped the product, case studies reducing record turnaround from 15 days to 48 hours and redeploying ...
Vic and Marcus talk with Carlos A. Rodriguez, MBA, CEO and Co-Founder of Humanate Inc., about his career from Bain and health system leadership to building a $40M SaaS startup and now launching Humanate, a company focused on agentic AI for healthcare. The discussion covers the origin of “Cassie,” a multilingual AI clinical concierge with ASL support, how staffing shortages and labor inflation shaped the product, case studies reducing record turnaround from 15 days to 48 hours and redeploying staff, a pricing model at half the cost of a human role with higher productivity, capturing patient-experience data at checkout, strict data privacy practices with no data storage and EMR-only transfer, NVIDIA’s partnership and vision for AI agents replacing apps, and the 60-day trial approach accelerating adoption across clinics and health systems.
Every week, healthcare VCs and Jumpstart Health Investors co-founders Vic Gatto and Marcus Whitney review and unpack the happenings in US Healthcare, finance, technology and policy. With a firm belief that our healthcare system is doomed without entrepreneurship, they work through the mud to find the jewels, highlight headwinds and tailwinds, and bring on the smartest guests to fill in the gaps.
If you enjoy this content, please take a moment to rate and review it.
Your feedback will greatly impact our ability to reach more people.
Thank you.
Okay.
Welcome everyone to Health Further Today we have another great guest episode, as you know, been really digging into ai.
Uh, where are we getting traction in the healthcare space using AI to really make a difference?
And today I have Carlos Rodriguez here from Uni Digital.
Carlos, thanks for doing this.
Yeah, thank you Vic.
I'm so looking forward to this.
I, I, you know, thank you for inviting me and I look forward to the session.
Yeah, it should be fun.
I, I wanna do, wanna disclose, uh, upfront we, my fund, my, my venture fund Jumpstart does have a position, equity position in Carlos's Company, human aid digital.
Um, but he is really one of the leaders in AI and healthcare, so we wanted to, to bring him to you.
So, Carlos, as we get into this, maybe give the audience a little bit about your background as, as a place to start, sort of how you came to this wild world of, of healthcare ai.
Uh, I think from a much more traditional healthcare background.
Yeah.
Yeah.
So I always start, you know, I was born at a very young age.
Yeah.
Uh, but no, my, my background educational wise is I'm an engineer, uh, picked up a a an MBA afterwards.
And, and really I kind of cut my teeth with a, a Bain and company and that was very heavy in m and a in telecommunications.
And this is back in the, uh, kind of early nineties and, and, and 2000 and, and Vic, it was just amazing.
You know, having worked for, for, for Bain and Company that really taught me a lot.
But one of the things they taught me is, is I, I was in charge of mergers and acquisitions and my specialty was the cost of goods sold.
So when we're looking at buying a company, if it had more than 20% labor content, we either said we need to automate it or we need to offshore it.
Right.
So that was the principle that yeah, that, that the labor is, was such a big part of our call.
So you needed to manage it.
Well turn the tables a few years back.
My wife gets diagnosed with stage three breast cancer and, uh, you know, and, and, and you know, being an engineer, I'm trying to figure it out and.
And a lot of it was because, you know, there was no transparency when it came to mammographies.
Back then it was the days with, with x-rays and it was very difficult to carry that x-ray from doctor to doctor.
Mm-hmm.
You know, before portals.
Yeah.
Uh, therefore whenever we went to a new radiologist, because we lived in a new town.
You know, they would not get the baseline mammogram, which is how you catch, uh, breast cancer because it's really the mutation of healthy tissue as you age.
You know, it gets unhealthier and that's what you get.
So when, when that happened, I said, you know what, I, I would like to change this.
So my first startup, you know, that sold at 25 x through private equity was coming up with the first cloud-based all digital release of information company when it came to medical records.
And my goal was to get rid of faxes, you know?
Um, yeah.
Out of, uh, you know, we're still working on that today, 10, I know.
20 years later.
We're still working on that today.
And, and what's funny, Vic, is, is, is I would go to Office Depot and OfficeMax and I couldn't no longer find, you know, faxes for our employees to, to purchase.
Mm-hmm.
Yeah.
And they would say, Mr. Rodriguez, the only two industries that still use faxes is flower shops and healthcare.
You know, so surprised to our flower shops.
I didn't know they were, uh, a long time.
Yeah, because they, because they get a lot of orders.
This is before, you know, you know, before people would do it, you know, digitally.
So we kind of came up with an all digital platform and it just grew like crazy.
Four employees to 400.
We're able to get UCSF, use the Kentucky Medical Center, Texas Tech, and, and it really grew, grew a lot.
So that's kind of, kind of, I cut my teeth, uh, inside of healthcare.
But to answer y your, your question, so now I, I'm in healthcare, you know, working in the Texas Medical Center and I've become the COO of, of Baylor Healthcare, and I find out that 70% of my cost is labor.
Um, and, and I'm trying to figure out, you know, how do you do that?
Because I was good at as cutting supplies and negotiating with vendors, but really, you know, how do you deal with labor?
The fact that in healthcare for every one doctor in the hospital, you have eight to staff, eight to 10 staff members.
For every doctor in a clinic you have one to four.
So it really, and it's hard to even find the people.
Nevermind, uh.
Cut the cost.
You can't even find them at any cost really, typically.
And, and, and, and that's really the realization, right?
So once we hit COVID, uh, with the startup company, it was so difficult to find people.
Um, and a lot of my peers were making the mistake.
Of, of using the government's money to just throw it at people, to try to retain them.
And, and it wasn't done in, in the way of bonuses when, when the money ran out.
So, so you inflated.
Labor by a good 20 or 30%.
So I'll give you the example.
Uh, most people in the front office, medical receptionist, uh, people that you know that do scheduling and continuity of care, you know, they're probably making 15 to $17 an hour.
Now you can't hire anyone over, you know, under than $20 an hour.
And even then, uh, they still have to pass, uh, the drug test.
They still have to, to, to, to be able to be fully immunized.
And it does become a constraint of labor.
Mm-hmm.
Yeah.
Yeah.
And so, uh, how did you, what was the origin story for Human Eight?
How did you think to do so your company is called Human Eight Digital.
How'd you come up with the idea, or where did it start?
Yeah, so as I started my company, I moved from Houston to a college town, which is a college station the home of, of 80,000 students at, at Texas a and m University.
And I started going to my primary care doctor and, and Vic in my 12 years in this town, about 200,000 people.
The average wait time to see my primary care doctor is now 22 weeks.
Yeah.
That's crazy.
22 weeks.
But it's true everywhere.
It's true in a lot of parts of the country.
Yeah.
And when I'm in front of my doctor, I'm with him less than 10 minutes and half the time he has his back towards me.
Yeah.
As, as he's typing my answers.
And, and, and I just figured there has to be a, a better way because, you know, think about that.
If you can't see your primary care in 22 weeks, who's the person that really knows you?
Then you almost, when you get sick, have to go to a walk-in clinic or a minor emergency.
Mm-hmm.
And honestly, those folks might not have your records.
They might not know your history.
Right.
And it is just not the way to do it.
And then also the inconvenience that certain folks need their medicines.
You know, refilled, but the doctor hasn't seen you in a while, so now he can prescribe them, right?
Yeah.
Um, so there's starting to be a huge imbalance between access to healthcare and then supply.
So I, I really thought that that automation was the way, you know, why not automate it?
You kind of see that in the banking world.
You see that.
In the airline world.
So I kind of felt that it was the perfect one.
And, and it's kind of funny 'cause we, we looked@indeed.com and found out that Brian College Station was trying to hire 54 medical receptionists.
Uh, they all had to be bilingual.
And I, I went and talked to a couple of professors at a and m and said, surely we can come up with a way that we can do self-registration without a human.
So that was really kind of what put us on the map, is the fact that we came up with a, the first medical receptionist that can speak a hundred languages.
And you can, and you can check in using your phone.
So, so that was the first, uh, concept.
Let, let's try to do a medical receptionist through, I guess through ai, is that right?
Um, where you interact with, um, an AI and it, it does the same thing the receptionist would do.
And, and you said a hundred languages.
A hundred languages.
And then, uh, we're, we're also, uh, the only company that has been certified the, the American Sign Language.
So we have an actual avatar that I'll show here in a second that, that she understands sign language.
So kind of our experience is, is you go up to, you know, to a screen or, or, or a computer or an iPad, uh, Cassie will greet you and say, Hey, I'm Cassie, the clinical care concierge.
Once you say your name, she remembers your name and starts speaking the language that you speak.
So even if you bring your, your mother who might be pakistanian or your mother, that might be Guatemalan, you know, Cassie is able to speak to each one of, of your family by their native language and, and that is huge when it comes to filling forms that have to be accurate.
Um, and that is huge also, when you're describing what's wrong with you, or in our case, we go ahead and also do a health and physical, if that's what the doctor wants us to do before, I mean, it makes sense that the Cassie doesn't get, she doesn't call in sick and she doesn't get impatient.
She doesn't have a bad day.
So you're live with a couple customers.
I know.
Do you have a case study or, or like a way to describe how this affects the practice?
Yeah, so I read a statistic the other day that, that just kind of floored me and, and that is that I remember that Americans, we hated our cable company because they never answered our calls and we needed them.
Mm-hmm.
Well, now healthcare has to pass cable companies.
Because one out of every three phone calls that you make to your doctor or healthcare provider does not get answered.
And, and, and the reason is that they don't have a digital access, right?
Uh, yeah.
And you have to call by phone and even to make an appointment, you, you get bounced around three, four times because it might be that a different doctor has a different scheduler.
Right, right.
And, and the person that you talk to has not the authority to book Dr. Smith on Wednesdays at right.
At at at two.
Um, so that's kind of what we did is, is we said, okay, so it's not just when I walk in.
It is also, I could, I could phone in or even I guess, online and, and Cassie would, would help me.
Yeah, so I'm about to show you a, a couple, you know, or, or you know, what we did?
It was our first customer, very bus, very busy orthopedic sports medicine practice, about 80 orthopedic doctors in the suburb off of Dallas called Frisco.
Um, and when I spoke to the CEO, uh, he was concerned that he was getting bad ratings.
From patients when it came to receiving their medical records on time, their imaging studies such as x-rays.
Uh, and then also one thing that that orthopedic doctors do is they have to fill a disability form, uh, for you to refile it with your insurance to get benefits.
So whether you break a leg playing pickle ball, you get hurt at work and you do workman's comp or you get hit by by SMI, all these things generates an FMLA form or a short term FMLA form.
Well, the government mandates that this form, you can only get reimbursed $25.
So it was very hard to get doctors and nurses to sit down out of a busy schedule.
And yeah, so the patient needs this form.
But the, it's not reimbursed at a significant level, so it takes them more effort than it's really worth to do.
But maybe, maybe you're gonna tell me, Cassie can do it pretty quickly.
Absolutely.
Because this is the beauty in, in oh eight, in oh nine during the Reinvestment Act.
Uh, you know, the government came up with a requirement that every medical record had to be meaningful use, and there was like 14 criteria.
Mm-hmm.
And the, and one of those 14 was it had to be machine readable without needing a human access.
Now people didn't take advantage of that.
Now Epic sure did.
Uh, and Cerner tried.
Uh, but you know, we're using that to our advantage, especially now that the inter portability and information blocking rule has come about, right?
So now I can access a medical record as long as I have permission from the patient or the patient's representative if they're underage or, or feel the HIPAA formm.
So check this out.
I now can go into the EMR Cassie because she's AI and really copies or scrapes all the DRG and CPT codes, fills the form, sends it to the human in the loop because we are advocates of having a human in the loop with ai.
That person reviews it and then slides it into the workflow for release.
So we're, so now it's $25 for a three minute review, as opposed to $25 for, I don't know, 40 minute to put it all together.
Correct.
So let me, let me share what that look like.
Yeah.
So what's the case?
So there's an orthopedic practice, you have a, something you wanna show.
That's correct.
Uh, so again, uh, the CEO has been very kind to us to be able to use them.
So this is the Carroll Clinic, uh, in Dallas.
Very busy from eight to five.
However, however, most of our healthcare happens around 24 7.
So, especially now with two, two members of the family working, we come home at five and that's when we really start thinking about, Hey, little Johnny needs a physical, or I need to go get this, I gotta get this, I gotta get this form, I gotta get this FLA form from my, yeah.
Correct.
But by that time, everything is closed.
Right?
Um, so the staff was also overwhelmed that they couldn't take the amount of phone calls from eight to five, and they just could not hire and keep enough staff members.
And then in this case, they had a huge bilingual issue that, that not all their members were bilingual.
So now you had to wait.
You know, for the Spanish speaking person to get off the phone.
And, and that was not so kind of what we did, which is kind of cool.
We said, you know what, we can do this through Cassie.
Cassie is, uh, an agenda.
Yeah.
So that's Cassie right there.
She has an actual, uh.
I don't know.
Not a body, but like she has, she has a presence.
Yeah, so, so somebody has a presence there.
Yeah.
So you're gonna see her in action, right?
So what we did is, is she is live on the now website of the clinic when it comes to patient service, medical records, disability.
And I'm gonna show you that in a second so you can hear or speak, right?
You can add as many languages as you want.
They chose Spanish because of the Texas demographics.
And this is what's kind of cool is.
I always love to sell stuff that has an immediate ROI return of investment.
Right?
It's very difficult to go to a hospital CFO, right?
Because he gets sold every day by By doctors.
Yeah.
And his margins are tight and cash flows tight.
Worse, so he, I mean, he has to be careful with what he spends money on.
Yeah.
You know, so the average hospital is three to 5% margins.
Um, and those expensive toys like a da Vinci robot, you know, you know, therefore the doctor.
So what we, you know, so how we sell and price our product is, is, is very easy to understand from release of, you know, from a return of information.
So the first thing we did is we were able to reduce.
The lag time, it took a patient to receive it from 15 days to 48 hours, and then we axle activated their cell phone.
So now they get a text message on their cell phone saying Your records are ready.
We give the the URL code and they can download them on their computer, and that has been a huge.
Uh, Sater for customers.
Yeah.
You're sort of meeting them, well, not sort of, you're meeting them where they live.
They're everyone's on their cell phone texting each other.
So now their doctor is, is right in that band too.
Yeah, Vic.
And, and, and we look at the IP address of, of the device that pings us because we use that as for transfer of custody in case people wanna know, hey, who's viewed my medical record.
And we're finding out that 80% of people that are accessing.
Our application is done by a smartphone, uh, and, and it's 80 year olds, right?
It's, it's not, you know, 80 year olds to, to, you know, to 20 year olds.
And then this is the beauty of it, you know, from a value proposition.
Uh, right now the average healthcare worker, it's about$70,000 on a staff level.
Uh, to healthcare, and that is about 20 to $22 an hour.
Unfortunately, benefits are very high in healthcare, they're close to 30%, so 70,000 is kind of what we would average, or I would average when I was a CFO.
So in this office alone, we've been able to redeploy, not, you know, not get rid of because they need 'em in different places, but we've been able to redeploy 1.5 FTEs, which is about $130,000 a year.
And that's from day one, right?
So the moment that you turn on.
Our app, because it's SaaS based, you're saving money and, and what's kind of cool is this is not money that needs to be approved by the CFO.
The CFO has already approved the budget as part of the labor.
It's just that that manager in that office cannot keep labor.
Right?
So a lot of times we kind of go into the office manager and say, Hey, you have five receptionist, job opens.
What if you hire four Cassie and leave one human in the loop and, and try it out?
Yeah.
Yeah.
So what is ca So it's saving, uh, a hundred thousand dollars a year or so.
Maybe 120.
I don't know.
I can't do the math.
And what, what does Cassie cost or how do you price it?
Yeah, so, so if, if you go, if you go to our website, which is very simple, it's human Nate, human nate.ai.
We have a pricing tab.
By an ag agentic, uh, uh, accountant named Joseph, right?
So he kind of looks like a very, so he'll walk me through the process on your website.
He'll walk me through the process.
So what, what, what, what Joseph would ask you is what are you paying that person per hour and what are your benefits?
And, and what we do is we model the pro forma and we start at 50% of what you pay that person.
Plus we guarantee 30% more productivity because we don't take breaks, we don't take lunches, and we don't have PTO.
So, you know, so what I always say is, is you can cut your labor in half with 30% more productivity is kind of like our tagline.
Yeah.
And that, that's a pretty easy pricing model.
I mean, whatever you're gonna pay your receptionist, uh, half of that is what Cassio would call roughly.
I mean, just for me to easily remember.
Yeah.
And, and, and now, and how and how so?
Can I listen to her?
How does she sound?
Yeah, yeah, yeah.
So, so I'm gonna do a, I'll always stay away from live demos, but let me do a live demo with an actual customer.
So you can see that, that, that this is not a, a, a, a gimmick, right?
Um, which is, which is kind of cool.
So I'm gonna take you to, to one of our customer's website.
I cannot complete the process because if not, I'll be generating.
Uh, a request to them.
Um, uh, but, but let me, let me quickly do that for you as you're doing that.
How long have you been, uh, when did you start the business and how long has it been, uh, in market where you actually had a product?
Yeah.
Uh, so we started the, we Incorporated, uh, as a Delaware C Corp in March of 2023.
Uh, so it's been a little bit over two years.
Uh, it took us a year to get the product out of the lab and, and make sure that it was safe and no bugs, and we introduced it in March of, of, of this year.
Uh, and it is, and it is, and it is generating quite a bit of of orders already.
So this is a website, uh, this is Carol Kling that I was speaking to before.
Mm-hmm.
Uh, they have 90 orthopedic doctors with their own ambulatory surgical center.
As, you know, orthopods make their money and the or, so they want to be as efficient as possible.
Therefore, they don't like filling forms.
Uh, you can see here how they just have doctors and doctors and doctors, right?
So they, somebody has to support those doctors.
So whenever you go to their website, all you gotta do is go to request medical records or FMLA.
You request the medical records.
And there's Cassie now.
Yeah.
They chose a different caster.
She's, yeah, she's dressed different or she actually has different, uh, hair too, maybe.
Yeah.
Yeah.
So it's a different character.
That's the one they choose.
You can choose any gender, any ethnicity.
We also have, uh, caricatures for pediatrics.
So then the first thing you say is, okay, I want English.
Right?
And you say, okay, uh, give me all my medical records.
Right?
You, you picked your language, English, and hear her out.
Hi, my name is Cassie.
Welcome to Carroll Clinic Medical Records.
I'm here to guide you through the process of ordering your medical records to place a request.
Yeah, it sounds like, it sounds like a hu It sounds like anyone really, yeah.
Yeah.
So it is kind of funny 'cause I, I read the reviews every morning and patients told me like, Hey, I feel like I'm talking to a real person.
Yeah, she is.
She is much better than filling forms.
Um, right.
So, for example, now I'll ask you for some basic information.
So then as she's collecting your information, it be, it becomes a chat GPT experience.
So you can either type it or you can say it, right?
And, and, mm-hmm.
And if you say it, she also does the verification, whether you're putting your date, uh, in, in, in, you know, in, in the proper format.
Mm-hmm.
Uh, we also ask you for your permission to take a picture of you.
So we validate your identity by comparing.
Your picture to the one that you gave the doctor, you know, uh, during registration.
Uh, we also capture your consent, uh, for releasing the medical records.
And afterwards we send you a text message saying, Hey, this is your tracking number, like FedEx, so you can log in and track the status of that.
And that is much better than having to call in the practice every day and say, Hey, do you have my request?
Right.
Uh, so we, we find that this is a great experience.
Uh, I would say that 80% of the people like it, uh, you know, and, and stay with the program.
We also let 'em opt out and, and try to call.
But 80% choose to stay with it.
And you always hit the wisecrack about, Hey, is Cassie single?
You know, because she looks so, so real, right?
Like, is Cassie single?
Um, but, but I would say that, that, uh, you know what, what I find meaningful is, is when a non-native speaker.
It's the first time I felt understood.
Right?
Yeah.
Right.
Especially, I mean, Spanish is, uh, I think reasonably prevalent across the healthcare industry, but once you go to any other language, it falls off dramatically.
And so if you happen to speak, uh, you said Pakistani earlier, that that's gonna be a lot of, uh, work to try to find someone in the healthcare industry that can speak with any level of fluency with a lot of the.
Communication and healthcare is fairly nuanced.
It's not like your restaurant travel, uh, vocabulary.
So you need to know much, much deeper.
You know, one of my classes I took was an international marketing, and, and it really taught you about words have different meanings, right?
Uh, so when, when they were, uh, marketing a, a vacuum cleaner, Hoover in Latin America, their tagline was like, nothing sucks like a Hoover.
Right.
You know?
Yeah.
And, and then in healthcare, you know, the, the campaign got milk got translated in Latin America to, are you lactating?
Now, you know, so, so think about that, right?
So, so those are the nuances, you know, in a different language that you know, that, that, hey, it hurts when I urinate.
You know, that might be something very personal that I need to tell you in my own native language.
And it's kind of cool, you know, I'm a native Spanish speaker.
I, you know, I learned English at 12 and what's interesting is Witt.
We didn't pick, you know, Spanish from Spain.
We didn't pick even the formal Spanish in Mexico.
Like gu.
We went with two.
Which is very informal because that's what Nicaragua, Guatemala, right?
Yeah.
So you have the ability to, to, to make sure that, that you're speaking to your audience and meeting them where they're at.
Mm-hmm.
Yeah.
Okay.
Now, I know you have a partnership with Nvidia.
What, what, what is involved in that?
How, I mean, that's a huge $4 trillion business.
What are they doing with, with, uh, little but growing, uh, humate.
Yeah, so I'm, I'm a, I'm a believer that, that there are times, you know, there are times in your life where, where the technology starts line up.
And, and let me give you a an example.
Two seminal events happen.
One was Steve Jobs said, Hey, this is the iPhone.
And at the same time, Jeff Bezos says, Hey, I'm introducing AWS and I, and I look at how disruptive those two things have been.
Mm-hmm.
And you know, everything has gone to an app.
Right?
So a couple of weeks ago I got invited to go to Google.
They were touring the Google Maps, right?
So this is where Google Maps occur.
And, and, and I start thinking, I remember traveling with my dad with an actual paper map.
Yeah.
And then on my own, I got a garment, GPS, and now I don't need that because I have an app.
You know, well, I, I really feel that, that the future is going to be agents on your phone, not apps anymore.
Right.
And, and that's kind of where we're getting to that, that you have your own AI agent that advocates on your behalf whenever you need to see a doctor, whenever you need to refill, refill, um, medicine, you know, from, from that standpoint.
Okay.
That's interesting.
So, so you're, right now you're putting the AI agent.
Uh, to represent the doctor, doctor's office, that orthopedic doctor.
Um, and so any patient that comes in the same ai, Cassie is, is its name, interacts, sort of on behalf of the medical group, but what you just said it, I hadn't really considered before that you might have an AI agent in your phone that sort of helps you with your own healthcare over time.
I'll give you an example, Vic, and, and again, respectfully, I'm not trying to be provocative, but, but the, the, the way, the way they sell us EMR in healthcare is that hey, you're gonna capture your clients and you're gonna have your own referral system.
Therefore no one leaves your system.
Right?
So, so, so let's say that, that I go to my doctor, uh, and he, and he or she sends me to do imaging.
Uh, you know, the EMR will try to schedule you within, within the healthcare system, and now I find out that I have a, a spot, you know, a, a cyst, a spot on my kidney, and now I need to go.
To, to a nephrologist or a kidney.
Why do I need to stay within that system if maybe the best nephrologist in my city is not part of that system?
Well, Cassie can can determine that, right?
And, and I'll give you an example.
You know, when, when I was, uh, at, at Baylor, uh, I looked at the cardiothoracic line, uh, because of Dr. Michael Lee Debei and de Cooley.
And I would find that, that some cardiothoracic doctors had 2% mortality and some had 0.2, which is huge disparity.
But you and I as patients never knew that.
Mm-hmm.
Why do I need to go to a doctor that has a higher mortality?
Right?
So, so I really think that these agents are going to advocate because they're wonderful search engines, right?
That can look through data and say, Hey, from what you told me, this is your best choice.
Yeah.
So I agree that's coming, but, but talk about what, what you're dealing with Nvidia right now.
Yeah.
What, what do you have in mind with 'em right now?
So, let's talk about Nvidia.
So the, the, the, the two things that technology has lined up is one is open AI chat, GPT.
That's a total experience of, of being able to search by voice.
The other one, you know, Vic, that is huge, is GPUs taking over CPUs as the most powerful part of a computer.
I, I remember when Texas instrument introduced a transistor on a semiconductor, and, and that was, you know, that was huge.
Right?
Well, this g you know, the combination of, of chat, GPT and Nvidia has been wonderful.
Nvidia took us on as an inception program because they want to, to, to incubate startups in the AI because we buy more stuff from them.
Mm-hmm.
Well, what are the kind of stuff that, that we buy?
We buy computational power because it's about $200 an hour to train.
You know, to, to train an agent in a, in a, in a GPU black wall, H 200.
So one is they're giving us steep discounts and credits to be able to use a computational power.
Second is they have, for us the best LLM.
So.
In NLM world, you know, large language models, size, size is not the most important thing, right?
So I tell my customer that we are 30% more efficient than chat GPT.
It doesn't mean we're better, it's just that we restrict casi to only pharma and healthcare terminology.
Yeah.
She can't talk about, uh, I don't know, f.
Astrophysics or something.
Yeah.
So, so if you ask her who's the better, the Yankees or the Red Sox, she's gonna say, I don't know what you're talking about.
Right, right.
But you ask her about, uh, a protein molecule and medicine or, you know, pharma, pharmacological products or anything like that.
Anything in healthcare, she really knows it.
Yeah.
And, and then the third part of what Nvidia has done for us, and we're very grateful, is that they put us as part of their venture capital ecosystem, uh, who was run by Howard, a Stanford grad.
Uh, so once a month I, I get coaching, uh, from Nvidia and they bring, uh, different venture capital firms and, and it is helped me a lot in terms of perfecting my pitch.
Yeah.
Now, how many startups do they work with?
Is it like a. Hundreds of startups that they're working with now.
It it, it is, but they're huge.
Right.
Uh, but I wanna show you something that, that I felt that we, it's almost like dating, you know, I kind of felt like we, we went past the first date, which is a startup.
Uh, and, and, and let me show you what, uh, what I mean by that, because this, this really has, has kind of put us on the map.
Um, so, um, they had called me, uh, Nvidia has an annual show.
In their headquarters in Santa Clara called GTC Show, and it's attended by 30,000 people a day for four days, and it sells out every day.
Um, and they had called me and said, Hey, for our startups, we offer you a a booth, but it's gonna be in the startup section.
You know, you have to go under the stairs and it's gonna be in the corner.
And I said, well, I'll take that right?
I nothing wrong with that.
And then they called back and they said, you know what Jensen Wong wants to do is pick five startups.
Five in, in computational power.
So we're pharmaceutical companies.
Five in, uh, physical ai, which is robotic surgery, which is a big one, and then five in ent.
So we were only one of five companies in the ent and we were right across the, the, the, you know, from them.
So we got a lot of traffic.
Uh, one of my highlights is I, I got to meet my hero, Jim Kramer from M-S-N-B-C.
Oh, nice.
You know, uh, uh, Kimberly Powell from Nvidia brought him over and he interviewed me, um, you know, from, from, from, from that standpoint.
Uh, but let me, let me show you, uh, kind of, uh, you know, kind of, kind of what, what happened there, which I think, I think you'll, you'll enjoy.
So this is an, this is an actual photo, uh, to be made looked like a Polaroid.
Uh, so there's Kimberly Powell who has been with Nvidia for 17 years.
You know, she's a, uh, MIT Stanford, Northwestern grad, you know, highly accomplished.
Yeah.
And this is March of 2025, so not too long ago.
Yeah.
So she does a keynote saying that the.
Agent ecosystem is gonna reach$1 trillion in healthcare, and that's with a T $1 trillion.
And then she announces they're partnering with eight companies.
And these are companies that we all know of.
McKenzie?
Deloitte, Accenture.
Yeah.
Yeah.
So if people, so people that are listening and not able to see it, Kimberly's on stage at NVIDIA's event, and behind her there are eight companies and yeah, it's, it's McKinsey, Deloitte, Accenture.
And then Carlos, what's the other one up there?
So, ate, right?
Ate, uh, ate and, and, um, since that happened in March, Vic, my phone has not stopped ringing.
Yeah, right.
Uh, we, we signed a contract with 50 orthopedic clinics all, all different.
Uh, we signed a contract with the largest healthcare system in South Carolina, 23 hospitals.
Um, and, and it just continues to, to gain momentum.
Yeah.
So I wanna talk to you about your sales and implementation process.
You have a really interesting approach to how you, you respond when someone calls you and says, Hey, can, what?
You know, I'd like to consider this for my.
My practice or my health system.
Um, talk about your, um, trial process.
Sure.
No, and, and we did that out of, out of, um, you know, out of experience and necessity.
You know, if, if you talk to any, any pharmaceutical rep, what, what they'll say is that the gestation period between your first meeting and when you get a contract.
Can be up to one year in a hospital system, maybe in a clinic is three to six months.
So it takes a long time to get vetted because it, it take you from department to department and you could go from it security to now procurement to legal.
Now the, the chief medical officer has to weigh in.
And, and a lot of it is you're kind of waiting on contracts.
So, so what I did is, is I, I said, let's go back to a handshake, right?
And, and I limit.
The, the risk and exposure to them.
So I say, Hey, I, you know, you don't have to sign a contract with me.
So what we do is we sign the scope of work, meaning this is the work that you're asking me to do.
You know, you want a medical receptionist, you want a, uh, you, you, you want a medical records, you want a billing specialist, you know, whatever.
Yeah.
More like clarity, kind of just clarity of what they're looking for.
Yeah.
And then what we do is, is, is we offer them a 60 day free trial.
Right where we go in there for a week, install the equipment, train them, and kind of babysit it for the first two, three days where patients are in.
And we take surveys and, and what I tell the, the, the decision maker is in 60 days I'm gonna come back.
And three things can happen.
One is you don't like it, and I'll take my, my Cassie back with me on, on, on the plane.
Second is, is you like it.
And, and I'll sign a contract or third.
You might like her so much that you want three more.
And one of the key KPIs for my company is the conversion from, you know, from trials to contract.
And right now we're running 95%.
Right.
So as long as we can keep those metrics Yeah, it's, it's a great way to kind of ease.
The concern about ai, what if people don't like it?
You know?
Is it only Yeah.
Well, and you get to really, I mean, it's like test driving a car.
You get to really try it.
You see people walk in and try it.
You try the internet, whatever, and, and then they know if they like it or not.
Yeah.
Yeah.
So I'll give you another example.
We're working with a healthcare system that sends an email for patient satisfaction scores.
And they were getting less than 4% responses.
Mm-hmm.
Right.
Yeah.
We put Cassie as a kiosk during checkout, and we've been able to increase that to 70% of people stop in and just spend less than a minute saying what they thought about their visit.
And it's not just rate from one to 10, it's things like, tell me who went out of the way to help you.
And, and people are surprised that it was not the doctor at times.
It, it was the medical receptionist that helped them.
Get to the right insurance card.
It was the valet person that, that, that got him a wheelchair.
Right?
Yeah.
So, so it's able to capture text by voice and, and I think people are learning a lot more about the experience of a patient when they go, you know, it takes about an hour for the whole experience to go to a doctor and it's less than 10 minutes with your doctor.
So we are saying, tell me about the other 15 minutes.
Right?
Were you able to find parking?
You know, were you able to pay, were you able to, you know, this?
And so I think, I think we're giving a more holistic view of, of the patient experience.
Yeah.
Excellent.
Well, so I have another question.
Um, a couple weeks ago in the roll up show, Marcus and I were talking about open ai, um, data.
Exposure.
So I don't know if you know, but they are, they hold all of their chats, uh, in history and, uh, Sam was, was St. All was, was on record.
He, he intentionally went on record telling people to be cautious because if you're in a lawsuit or whatever, it's discoverable that the law enforcement or the other side can, can get their chat records from OpenAI about you planning to do.
Learning how to break into a bank or whatever it is.
Um, and Marcus and I were talking on the show about a lot of people are sharing, uh, behavioral health things, physical health things, very intimate health details with their, uh, AI because it is very useful and they can't reach their doctor or whatever.
It's late at night and they wanna do it, whatever the various reasons are.
And that creates a, a pretty big.
Data security risk.
So how do you, you're handling a lot of healthcare data.
How do you manage this?
And Humate, do you have that same thing or have you solved it?
No, absolutely.
I, I think, you know, I think we have a very robust solution that, that, that can withstand, you know, all the cybersecurity and hacks that are happening.
So, so let me give you a couple of things.
First of all, we do not store any data any, so, so when you look at, you know, when you look at hipaa, which, which which governs privacy, right?
Uh, we do not store any data because, you know, it's almost like printing a piece of paper.
Once you print a piece of paper, it's a breach about to happen is somebody finds a piece of paper.
So we do not, so, so when you're doing the intake, you're talking to me, you get my history.
You fill out the electronic medical record in the doc's office or in the hospital.
But, uh, human aid is not itself holding any data.
So it's the same data risk that I would have in any health system.
My data's in the health system of course, but, but human aid's not holding it.
Yeah.
And we go even beyond that.
So, so check this out, Vic.
It's kind of cool.
So you go up to her if, if you choose to, to, to get it through a kiosk and, and she'll say, I'm Cassie.
Who are you?
I'll say, I'm Carlos.
She will always remember your name and call you by your name, Carlos.
But as soon as that inter interaction is stopped, we erase her memory, right?
We fully erase her memory because we don't want the next person to say, who was that and why were they here?
She knows that right?
She knows that, that, hey, I'm there because I have shingles.
Maybe I don't want people to know that I have shingles.
Yeah.
And does, does Cassie recognize my face or is she, uh, how does she know it's still me?
Yeah, so we're working on, on different identities right now.
What we do is there's a part of the medical record called a master patient index.
That's the most accurate file that has your information.
It's eight demographics, names, sex, age.
So what we do is we have an algorithm that lifts it from the federal id, whether the driver's license, and we've gotta match all eight criteria, right?
So we match the criteria in your medical record to the criteria in in, in that.
And what's kind of cool is because she is smart.
If she notices that you have a different address, she'll ask you, Hey, uh, which address is the most current?
And you say, oh, I just moved.
She'll congratulate you for moving and hope that you like it.
Same thing with getting divorced, changing your names, and we're able to then correct the information within the firewall.
Of the client's database.
So we don't collect any data.
We always transmit the data back inside the firewalls of that EMR, right?
Yeah.
So whatever.
Yeah.
So it's actually, um, maybe more secure than the human way, right?
Because the human.
Doesn't wipe his or her memory with every patient walking by.
They, they, they respect it and most of the time they don't do anything with it.
But if you wanted to be a bad actor, you certainly could.
And this is the same thing.
Cassie is electronically putting data into electronic medical record.
But then as, as soon as that interaction's over her memory is wiped, and she moves to the next one.
I'll give you an example.
So, so, so when I was in Houston, uh, one, one of our residents, unfortunately, uh, was shopping late for, for food.
You know, they're always working late.
And he got carjacked and got shot.
Uh, well then it became part of the community.
Hey, did you know a resident Joe got shot?
And people go, yeah, I, I know Joe.
Yeah, he got shot.
So, so people were talking about, uh, his medical condition and went as far because they were also residents to log into the EMR and see how Joe is doing.
Well, Cassie.
She might've seen the shooting, but she forgets about it, right?
Mm-hmm.
So she doesn't able to to say, Hey, what happened?
You know, uh, uh, so you can really, you know, uh, uh, do that in terms of erasing her, her, her memory.
Yeah.
Okay.
So tell me how the, uh, the 60 day trial has worked.
How, where are you?
Uh, you said something in South Carolina, you having good success, um, and then what kind of practices or what are the re what are the ways that people are using.
Uh, Cassie, or, or maybe you have different personalities too.
Yeah, yeah.
So, um, we're calling ourselves staffing as a service, almost like a play on words on SaaS, right?
So, so in some ways, when, when you look at our website, again, human aid.ai is you, you get to pick your character.
You tell us what kind of character you want.
And then you go to their resume or, or you go to their job description, they'll tell you what, what all, all he or she does.
And then you can also look at the resume and see where they're working.
So it's almost like hiring a candidate.
So you say, okay, well this is what she does, because that's a job description and this is where she's working.
So you can call and see how the referral, what, what does that mean?
Where she's working?
Like, uh, I guess Oh, where she's, where, where she's currently deployed at, at a current, yeah.
So like, uh, that.
Uh, was it Carol or something related Clinic Orthopedics.
So Cassie is working at Carroll Orthopedics doing A, B, C, and D as a receptionist.
Yeah.
Correct.
Yeah.
Um, uh, so, so in some ways right now, uh, we try to stay away from any credential jobs.
Uh, in full disclosure, I'm the only non-doctor in my family, meaning that I have two brothers that are doctors.
And I can tell you that, that if somebody goes into my brother's exam room and he's a Mayo trained surgeon and says, Dr.
Rodriguez, this is what's wrong with me and this is what I need.
He gets incensed, right?
Because he says, I went to school and now you know more than me because you, you googled.
You know Right.
The website.
So, so we stay away from, from medical decisions.
Now we're able to do them and maybe in the future, but we think a a, an entry point that is highly sought by doctors and nurses is to have a full staff.
Yeah, and, and they don't care if it's digital or humans and at times digital perform better because they show up every day.
And they have no bias.
They have no bias.
Yeah.
I mean, if I think about the, the doctor friends of mine that they have incredible turnover in their non-credentialed staff.
Almost always they have open job.
Open, job open.
I guess that's obvious.
They have o job openings.
Um, and so what you're saying is, let's let us help you fill that need and get your patients taken care of faster.
And, and it actually probably drives more throughput through their system.
They can see more patients, I would think.
Correct.
So for example, um, um, the, the university that you know that I'm next to.
Uh, we get our money from the state for rural healthcare, so we're supposed to stay on our lane of rural healthcare, which is a huge issue as you know, with rural hospitals and clinics.
So therefore, one of the hardest, you know, clinical lines to access rurally is adolescent.
Uh, psychiatry.
Uh, so now a and m has a telemedicine behavioral program that even teachers can call and say, Hey, little Johnny's acting a little bit differently.
Do you mind doing a 30 minute evaluation?
Well, they did 260,000 evals and ran out of doctors.
So we said, Hey, let's help a doctor and we'll conduct the health and physical.
So we, we get on the screen before the doctor.
Uh, we read from the medical record how you answer anxiety, depression, self harm, and so forth.
We ask you those questions, score them, and then send 'em to the doctor before they get on.
And what we find out is that people are more honest with Cassie than they are with a clipboard or a doctor.
So if you disclose that you had an alcohol or a drug problem, and we say, how's, how, how many drinks a day you're having?
We've followed up and find out that people are more honest with Cassie than they are with their provider.
Hmm, interesting.
Okay.
So now Carlos, give me a sense for where you think this is going.
Like, uh, both at Human Eight and for the industry overall.
It says on this slide that we're looking at, and Nvidia thinks it's a, it says a $1 trillion opportunity.
Seems incredibly large.
Where, where are we all headed with, with AI in healthcare?
What, what is your, you're close to this.
What do you see?
What should I be looking for?
What should listeners be looking for over the next, say, 12 to 24 months?
Yeah, great question.
Uh, I, I think AI is perfect for healthcare, uh, because it, it, it, it's almost like, you know, when Southwest came out, you know, they said, Hey, we'll, we'll give you better service at a cheaper price.
Right.
It's the same thing with ai.
It gives you better service at a lower price, right?
So you don't have to compromise one for the other.
So when it comes to things that patients need to access the, you know, their healthcare, whether it's a doctor's appointment, whether it's a refill, asking for.
You know, asking for a referral or on the other way from a payer, you know, think about, we have groups of people that all they do is answer denials.
80% of collections can be done through, through ai.
Right.
So I, I really think that that AI will be the first line of defense before they triage it or escalated.
To a very senior subject matter expert.
Right?
Yeah.
But you don't need 10 people.
You might only need two or three.
Yeah.
It's interesting.
I mean, hell, everyone, a lot of people are scared that there's gonna be job loss.
With ai, and I think unlike many industries in healthcare, we don't have enough people.
We, we, we have a shortage of nurses shortage, of doctors shortage of almost every, uh, role.
And so I don't think anyone in healthcare that I've talked to is worried about losing their job.
They just need help.
And so it could be really a, a, a good match where it's not gonna, you're not gonna lose.
Receptionists are not gonna lose their job.
We just need 10 times more receptionists to treat, take care of all these patients and get them the help they need.
Yeah.
So let me give you one more example.
I know we're getting short on time, but.
I, I, I was talking to my brother again who's a surgeon, and, and he was very frustrated because he gets paid based on DRG codes, which means that he gets paid based on his productivity.
And he says that the chairman sat him down and said, Joaquin, you need to slow down your, you know, your, your, you know, your operations that are not emergency operations, uh, you know, like colonoscopy and so forth, because we just don't have enough nurses.
Uh, and they're getting burned out and.
Travel nurses are now 150 bucks an hour, so now they're trying to, to go back to that.
So long story short is, so I said, Joaquin, what do you do with your nurses?
And he goes, well, before a, a patient comes in for an operation, I have to sit down my nurse and they have to do a preoperative screen.
A preoperative screen never changes.
It's always what medicines are you on?
If you're in any blood thinners, you need to get off of 'em because you know, you can get operated, you're gonna bleed to death.
Yeah.
And if there are exceptions that gets referred to the doctor normally.
Exactly.
Yeah.
So I, so I said, Joaquin, we can do that for you.
You know, ask all those questions, put 'em on a report and send it to your nurse and, and she will go from a 45 minute call to a 10 minute task.
And I can tell you that, that nurses prefer to be with a patient than, than to be in front of a computer typing.
Yeah.
Okay.
So, uh, that's pretty exciting.
What's the one thing that that's not figured out in your mind that we have to try to figure as an industry to, to unlock this?
Um, a a couple, you know, a a couple of things and, and, and to me it, it is more around what you say around privacy.
Right.
So we, we, on purpose, because of my previous experience with medical records and release information is I try not to keep data.
My concern is the last 20 years, the internet has grown by reselling our data.
Right.
And, and we gotta get away from that model that, you know, that that model, that, that, hey, if, if I don't get charged for the service, I, I, I, I am the product.
Right?
Right, right.
Um, so, you know, AI is gonna know a lot about you, but you gotta make sure that it works for you, not for anybody else.
And there's a, there's a state out there that is trying to pass the opt-in law.
You know, if you look at all of our apps.
It tends to be, hey, you know, read the fine print and you can opt out.
Yeah, it ought to be the opposite.
And that is that, that you need to opt in if, if you want your data to be, to be resold or shared with anybody.
Hmm.
Interesting.
Okay.
Excellent.
Well thanks for doing this.
Uh, how can listeners find out more?
I mentioned the website one more time.
I know you said it and then I'll put it in the show notes, but then.
Maybe do you have a LinkedIn or email or how do you like people to contact you?
Yeah.
Uh, so human aid AI is, is the best way.
Uh mm-hmm.
It, it has all of our contact information, so we'll be happy to schedule a demo with you.
Okay.
Uh, you know, from, from that standpoint, we do have a LinkedIn page.
Uh, but I would say that, that, that human aid, ai, AI will, will, will get to me within 48 hours.
Okay, so human, its H-U-M-A-N-A t.ai, is that correct?
Correct.
Yeah.
Yeah.
Okay.
And it's, it's not a word, but the, the invocation is to make something more human or more humanlike, right.
So, so we call Cassie the clinical care concierge.
Because we want her to be a concierge, and I don't know if any of of our viewers are in concierge medicine practices, but you know what I'm talking about, that if you have a concierge medicine, there's a lot more thorough follow up.
Yeah.
Yeah.
And kind of bringing that, uh, really in depth follow up, attention to detail and patient care to the broader industry is pretty, pretty interesting vision.
Yeah.
Yeah.
Yeah.
And, and what our CMO, we have a chief medical officer says, you know, Carlos, one of the hardest things about being a doctor is everything you have to remember up.
And he says, by, by having your technology, you know, I don't have to have everything up here.
And, and, and he sees it as a, as a great advantage that, that he can reference different things.
Yeah.
Excellent.
Well Carlos, thanks for doing this.
Super excited about what you're building with human aid.
See you appreciate it.
And uh, I'm here to answer any questions as a follow up.
Okay, great.
Thank you.
Take care.
Bye.