50% Documentation Time Savings? The AI Revolution Hitting Post-Acute Care
Featuring Alex Milani
About This Episode
Transcript
Tom Maxwell (00:09)
Welcome back to the Last Visit First podcast, episode number 16. Super happy to have today's guest, Dr. Alex Milani. He's the CEO of Steno Health. Alex is a physician entrepreneur focused on using artificial intelligence, and you guys have all heard me talk about artificial intelligence. It's really clinical intelligence, it's not artificial. reduce documentation burden, in clinicians and improve care delivery across the home base care.
We're going to talk about this in a lot of detail. Alex earned his MD and his MBA. So he's one of them really weird guys that has an MD and an MBA from Emory University. And he's the CEO and co-founder of Steno Health. So we're going to get into a lot of detail around the latest and greatest announcement that just came out in our industry around Steno Health and Home Care Home Base. So welcome to the show, Alex. Glad to have you.
Alex Milani (00:58)
Yeah, excited to be here, Tom.
Tom Maxwell (01:01)
That's great, that's great. Let's jump right in. We started a collaboration between Steno Health and Home Care Home Base. You guys just announced that. Tell the audience what this is about and what are we gonna do and why.
Alex Milani (01:12)
Yeah, no, it's something we're super pumped about. So essentially we're embedding some of our technology inside of the home care home base EHR in order to reduce that documentation burden and administrative burden these clinicians face. So in a few short months, this technology would be deployed at scale across the huge footprint home care home base has.
Tom Maxwell (01:32)
Yeah, so 195,000 nurses, chaplains, social workers, aides, PTs, OTs, STs are running around the country carrying an Android tablet and doing documentation. It's the number one thing that I hear complaints about through my consulting company is how long the documentation takes, how rigid it is, how complex it is. And you're the solution that's going to make this easier. How are you going to do it?
Alex Milani (01:57)
Yeah, so definitely I always ⁓ pick on Oasis, but we do HOPE, Hospice, like other forms as well, but I always tend to say Oasis for some reason, but it's definitely a highly scrutinized multiple choice test, I would say, that is rich with clinical information that takes a lot of time and focus for these clinicians to do. with our system, a clinician enters a patient's house,
They consent them to be recorded. Assuming that consent is approved, the clinician conducts their interaction with that patient as they normally would, just listening in the background. And from that interaction, we're able to actually fill in these items, whether it's OASIS, non-OASIS, agency-specific questions from that interaction. After all that, the clinician has a very unique debrief with AI to fill in those gaps and then guide them through that review process. The big thing.
that would want to highlight is this is definitely not an autopilot tool where it's, know, Hey, it replaces clinical judgment. It's a co-pilot. that clinician, would say intuition, you know, the thing they've been trained to do for so many years is embedded in the system where they, you know, fact check the AI proofread everything. And so it's just a co-pilot along their journey.
Tom Maxwell (03:11)
And so the goal is to reduce clinical documentation time or what we call pajama documenting. Shout out to Tracy ⁓ Dent, Elvar at AccentCare. She's the one that named it pajama documenting. But essentially what happens is a nurse goes into the home, does the basic physical assessment information, grabs the meds, allergies and those kinds of things.
They leave the home, they go home and document at night in their pajamas, is what Tracy calls it. And so our goal is to have clinicians documented in the home, quicker, faster, better, and reduce that time it takes to do pajama documentation. You guys have been using this tool for quite some time. I know that there's quite a few customers that are piloting it right now. What's the response rate? What are people saying?
Alex Milani (03:48)
Yeah, so it's been two years of development, been a long haul. Response rate has been pretty amazing across the board. The big thing I would argue is this is one of the bigger transitions in healthcare from, always, maybe it's hyperbole, but went from paper to EMR, now EMR to something that's different, AI. it takes a few weeks to get used to. This is a change in workflow and a change of...
how clinicians operate in the field. But usually after week three, after week four, we can get up to around 50 % time savings and documentation time, more focus on that patient. And this is a lagging indicator, but actually more accurate documentation as well, which is very, important, not from an audit perspective, but also for patient health.
Tom Maxwell (04:32)
That's awesome. clinician gets the device, they click a button, it listens to the conversation. We have to retrain those clinicians on how to make sure that they're talking throughout that visit. Nana, how's your pain? How's your back? Your back's hurting. How, what makes it hurt? When does it hurt?
relieves the pain? What time does it start? What does it feel like in the morning? What does it feel like at night? All those kind of things are conversations that you're having with that patient anyway. And then the device is essentially grabbing that information and saying, hey, based on what we've heard, based on the interaction between the patient and the clinician, here's how we think that this documentation should read or look. And then the clinician then gets that, they read it, they quickly review it and say, yeah, you're absolutely right. You don't have to go type 7,000 characters out on an Android tablet.
does that typing for me. By the way, it doesn't make spelling errors, it doesn't make errors, all those kind of things. It's amazing technology. again, I call it clinical intelligence versus artificial. Artificial just sounds like such a terrible word. It's not artificial, it's real clinical intelligence.
And it does a great job. I've seen it in work. I've seen it in action. I'm a huge fan of what you guys are building and what you're doing. By the way, this is what the industry needs really bad. I'm glad Home Care Home Base is endorsed and signed off on this to really get after it. mean, it's, I mean, we developed Point Care in 2005. It's still the same as it was in 2005. Now, sure, they have updated it. They vet the, know, A, C, E, over D and C and B, but the click, click, click through all the different physical assessment questions is gone. It's history. Now we're going to do really good, high efficient clinical.
documentation. Talk about some of the validations and things like that. You and I have talked about this a little bit in the past. And by the way, guys, I spent a lot of time with Alex on this. mean, he's one of his biggest fans and he's the one that stretches my brain the most when we start talking about things like this because I'm so excited about it. clinical validation. So I'm a patient, I'm on Coumadin, I need a PTI and R. In home care home base, you schedule a buddy code. The buddy code tells you when you need to go grab that PTI and R every two weeks.
Alex Milani (06:07)
Yeah.
Mm-hmm
Tom Maxwell (06:30)
that are clinical
understand what this means. The artificial intelligence should be able to do that or the clinical intelligence should be able to tell you, hey, it's been two weeks since you had a PTI and R, go grab another one, right?
Alex Milani (06:40)
Yeah, so basically you're going to the other tip of the iceberg moment. So right now, like phase one essentially is reducing that, you know, pajama time documentation after hours. But since we're uniquely in the home with that clinician, we're somewhat seeing the patient in a way. And we also have some sort of intelligence on the back end. We can actually service these insights at the clinician, the back office as relates to the patient care.
that's just really like the tip of the iceberg, a good example of what you said. Other ones that are kind of low hanging fruit as well are comparing this follow-up or this discharge with that start of care. How has the patient progressed over time? What interventions have been given? But essentially the very first step as relates to AI and home health or healthcare in general is...
The wedge in the door, if you will, is kind of that ambient scribe for reducing documentation. But the next level is, know, clinician intelligence, empowering that clinician as much as possible to, you know, reduce readmissions, improve that patient experience and their own experience as well.
Tom Maxwell (07:45)
Yeah, so Stephanie Chaney was the original nurse that developed home care home base with us way back in the day. We're going back to 2003, four, five, six when Scott Beard and I were April and all of us were with Stephanie and she used to call it the little angel on the shoulder. So the little angel sitting on the nurse's shoulder going, Hey, make sure you remember to do this. Don't forget that. You know, those kinds of things. I think with this clinical intelligence that you have now with this, um, the ambient listening stuff is going to make to make that angel more powerful.
So think about a CHF patient, congestive heart failure patient that needs, you our goal is to get a weight on every one of those patients. Cause if they start, you know, retaining water, they have a problem. They're going to essentially drown or, you know, have their lungs fill up with fluids. And so if we can always remind that clinician, Hey, this patient's a CHF patient. Don't forget to get a weight. wait, you can't get a weight and then it can't get out of bed and get on the scale. Then what about let's do the arm circumference or let's measure the calves and make sure that we're not, you know, retaining fluids. So those are the kinds of things that are going to make it really cool, but also make it really fast documentation.
Alex Milani (08:13)
Yeah.
one thing also that's so unique in home health or just in post-acute in general is you're in that patient's home, you actually have the most up-to-date med profile on that patient because you were literally in their cabinet hand grabbing the meds or you're taking the lacinopril or whatever. And there's a layer of intelligence that not only is empowering clinicians but also
we theoretically can empower the overriding physician or NP depending on the state. Like for example, 40 % of heart failure patients are on GDMT, know, guideline direct medical therapy. We have the most up-to-date meds, value-based care is now in home health and across the board. The question is, you know, hey, should we be surfacing this insights to that PCP as well? So like, it's just such as interesting avenue where...
or not only providing care for this patient and helping with the clinician on that home health side, but longitudinally we actually can reduce readmissions by servicing to their physician on the acute side and so forth. And so that's a level we're also exploring as well with our chief medical officer. But it's just kind of interesting how home health is this kind of linchpin or spider web, if you will, of healthcare where things normally have been forgotten, but you actually can do a lot more than you think.
Tom Maxwell (09:52)
That's amazing. And so all the social determinants is another area that I think you guys can look at falls. Does Nana have running water? Does she have power? Does she have electricity? Does she have heat and the so.
heat in the winter, air conditioning in the summer. Is she lonely, scared, bored, sad? All those kind of things are things that we can find out really easily from the, just from being in the home and doing that documentation. And so a nurse that walks up to that medicine cabinet, go back to that story, that's pretty good, walks up to the medicine cabinet and she talks through, here's all the medications you're on, here's the dose, here's the SIG, here's the instructions. Is the intelligence smart enough to pick all that up and then go update the med profile?
Alex Milani (10:05)
Yeah.
Yeah.
Yeah, so there's three different ways. So meds is everyone's favorite when we demo or, you know, launch a pilot or launch the platform. It's usually the one that gets the most jaw drops. So there's three ways to kind of get meds into our system. One is actually from the prior record. So I'll just pick on a start of care. based on that discharge summary, we pre-populate that start of care with the meds. So a lot of times they're actually removing meds as opposed to adding meds, which just takes a lot, you know, saves a lot of time. The fan favorite that most people do
is either line up six or seven pill bottles, take a picture if the human eye can see it, they eye can see it, and it extracts that from there. Or the other fan favorite, if you will, is home health there's this mandatory requirement, you have to hand them that piece of paper of their meds on that start of care, you can just simply take a picture of that as well and extract it. But we're also able to pull it from the transcript as well, but most people like the photography, funny enough.
Tom Maxwell (11:22)
That's awesome. So way back when I was at Homecare Homebase, we were the first ones to do the SureScripts medication request. so, one of the big problems with it is,
When I discharge from the hospital, have way more meds than I'm probably going to be taking. so, and they're not up to date. The SureScripts meds weren't up to date because in order for SureScripts to have the medication on the record, when we do the request, they have to process the payment. So it all gets at what they call a switch in pharmacy. And so it has to process through the switch in order for it to be on your record. And so what we heard mostly was it was inaccurate because the patient just got out the hospital, they had five meds to go get filled. Their daughter goes and drops them off at the pharmacy and goes and picks them up the
next
day and so they haven't processed through the switch so the nurses always having to discharge or clone medications or add new meds in. But what you're saying is you're able to do it from a real-time visualization of those, take a picture, drop them in. My study was an expert in point care so Donna Winchester was the expert that we used way back in the day. By the way, Donna is probably the best point care user in the world because she demoed it for 20 years for me.
but it would take her two minutes per medication to load a medication in. And the average patient in home health is around somewhere between, average Americans on four medications, believe it or not. I heard that this morning on a podcast, I was just too, but the average home health patient's on 10 to 11 medications. So just the two minutes it takes, that's 22 to 30 minutes that we're dealing with on the average nurse just filling out meds. And you're saying you can do that within a couple of probably seconds, maybe a minute?
Alex Milani (12:32)
Yeah.
Yeah,
it's, ⁓ you know, the seconds I would say, I would say devil's advocate a minute or two, yeah.
Tom Maxwell (13:00)
That's crazy. Okay, that's exciting. If you're listening to this podcast, you be looking at that. That's going to save your nurses a ton and ton and ton of time. Holy cow, this is pretty cool. I'm excited. All right, Alex, let's talk about you a little bit. So you're a MD and an MBA, which those are like, you're not supposed to be those two. Most doctors are not MBAs. So what made you change? made you, tell me about your thought process and your history there.
Alex Milani (13:27)
not the most common path, especially not to be practicing or obviously doing more startup-y stuff. I always kind of wanted to be a doctor ever since I was like eight years old. So if anyone asked me as a kid, what are you going to be? I'm going be a doctor.
And this is much because I looked up to my father who's a physician and a cardiologist. So after undergrad, went straight to med school. Emory really loved my time there, amazing institution. But I didn't, I guess.
I didn't get that addiction kind of feeling that I wanted in my career, if that makes sense. So I love treating patients, but it didn't really, that switch didn't flip as much as I wanted to. So I did an MBA actually because I panicked. So in between third and fourth year, did an MBA at Emory. And I was able to actually convince the Dean of the med school to let me take one more year off this MBA and work at a startup. during this time, I...
know, I thought I would like working at a startup. know, I heard about them listening to the podcasts about startups thought it could be cool. So I emailed probably 50 different startups a day through their customer support channel, trying to find like an internship and a job. And one was lovely enough to hire me. And so I took a year off work there and I was like, you know, fish and water kind of situation.
So after I did that, I knew I wanted to do something very different. then after graduating, obviously contemplated, know, residency or going directly into startup would have been internal medicine if I had a pick. you know, went straight to startup, worked at a variety of different startups, got into AI and healthcare, and then started this own company.
Tom Maxwell (15:00)
Well, that's So do you still practice medicine at all or are you done with that?
Alex Milani (15:04)
No, So, rightfully so in the US, if you want to prescribe and see patients you need to do residency. I would say I have more of a PhD in medicine, healthcare delivery, than practicing medicine. So, no, cannot, don't see patients, don't prescribe or anything like that.
Tom Maxwell (15:22)
That's good, but I'm glad you're doing this, because I think having that healthcare background while you're trying to build technology for nurses in the field, I you have enough experience to understand the language of healthcare and the language of medicine, so that's awesome.
All Back to Steno. What's different? So there's 25 different companies that are out there right now that are all talking about AI and ambient listening and automation and all that. What's different about Steno?
on the outside and then also what's going to be really different about Steno and Home Care Homebase together. That's a super interesting one for me.
Alex Milani (15:54)
Yeah, definitely more and more saturated space, I would say, by the day. There's almost a scribe vendor that gets released every lunch. So it's very interesting space to be I guess there's a wide variety of ways to answer this question. relates to our integration with Home Care Home Base. This will be the largest and only direct.
Tom Maxwell (16:01)
You
That's right.
Alex Milani (16:18)
true integrated system its kind in home health and hospice. There's many publications. The most famous one is with Kaiser as it relates to physicians and nurse practitioners using scrappy technology. The ones that actually adopted and use it were the ones that were embedded in the system, not two separate platforms. And that was for SoapNotes. So something where it's easy to copy and paste. But for home health, this is a very different beast.
if it's not integrated, it's hard to really launch at scale. There's a reason why CenterWell or Amedisys or some of these larger ones don't really have a solution as well as it's described. They were waiting for Home Care Home Base to build this, rightfully so, they wanted to integrate it. So that's the big tail in terms of our differentiating going forward. What got us to this point is a variety of different ways of how we
solve this problem. Truly, guess our company thesis has been scribing as a wedge, it's not the end state. So we kind of talked about earlier about, you know, surfacing clinical intelligence to the clinician. We also do things by empowering the patient by sending them a patient tailored note that is tailored to the health literacy. a wide variety of things we do from this tiny wedge to not only help the clinician, the patient, other means as well.
So kind of what got us on the map and on, I guess, home care home base's is radar. And then it just so happened that our next few years kind of partnership looked, our roadmaps were almost identical. And so it was easy sell on both sides to kind of get this partner solidified and now into the hands of the clinician.
Tom Maxwell (19:03)
As you think about advancing and going, not that I want to your feature creep on your product roadmap, but lots and lots and lots of patients now are requiring, they're not just English speaking, they're Spanish, they're Chinese, they're Russian, they're all the different other languages. That's a really interesting use case for, in home care home base, they attach the spoken language of the patient.
Alex Milani (19:10)
Yeah.
Tom Maxwell (19:28)
Primary spoken language, think is what's called in the clinical assessment. Why can't we take that and convert it? And so you guys could easily create that patient literacy documentation and I mean, Spanish, Chinese, Korean, whatever it is, whatever the spoken language is. you already do that, that's awesome.
Alex Milani (19:41)
We do.
Yeah, we're
ready do that. languages, there's always a, during the, I guess, demo phase or piloting phase, there's always a few set of questions and that's one of them. Languages, we support over 80 languages. 40 of those have a 95 % accuracy of an award error rate. But in addition to that, you know, the communication actually back to the patient is in their language or the loved one's language. So it doesn't have to be English and that's, you know,
plenty of studies that also indicate this folks in the US whose primary language for whatever is Spanish or Ukrainian, tend to get worse health care because of lack of communication. So we're trying to solve some of those barriers, specifically in home health. And if they got an English documented, they're not going to know what's happening. So we obviously tailor to their reading grade level and their language, of course.
Tom Maxwell (20:37)
That is awesome. I didn't know you already doing that. That's cool. And there's so much more you can do with that as well. ⁓ Just on the medication list, mean if you get a prescription medication that's in English and you don't read English, you read Spanish or you read Russian, how are you supposed to take your medications appropriately? It's a big problem I think, so ⁓ that's really
Alex Milani (20:42)
Yeah.
Tom Maxwell (20:57)
What are the things that are real popular when you do your demos?
Alex Milani (20:59)
Yeah, I'll give you the popular things and I also want to talk about the skeptical things. I think that's also very, very important. The popular one, I would say easily is the meds. I think you nailed it. know, is everyone's, it's their bane of the existence to grab that giant sack of pills, read them out, write them down. It takes like an hour. that's usually the jaw drop moment.
Tom Maxwell (21:06)
Sure.
Alex Milani (21:22)
But also what we do with the records before the visit happens is something that we're very keen on exploring even further, but we already do stuff with it. So much a clinician is supposed to pre-chart on their patients before the visit, our AI does the same. This is to give the AI an idea of what the patient looks like, risks, some socioeconomic factors, mental state of the patient going into it, just to know if their voice should, how much weight should their voice be taken in?
For example, if they're nonverbal, you obviously, you know that the loved one's going to be speaking. If they're demented, maybe they're maybe not in the right state of mind. So the clinician's voice matters more. It's pre-charging essentially. But we also service that to the clinician as well. Like these referral packets, these home health agencies and these clinicians get are, can be up to 300 pages. It's hard for them to find time in the day to read through each and every patient before the next encounter much than, know, yeah.
Tom Maxwell (22:14)
They don't do it.
Alex Milani (22:16)
It's also hard for them because they're documenting every night for four hours and making dinner for their kids. it's, I completely understand why. So we're trying to streamline that as well from the prior documentation. So the two big wows in addition to the scribing, because everyone knows about scribing, are the pre-records and then the meds as well. But the things that I would always want to highlight is, which I found the most interesting actually early on was
When we onboarded with clinicians, their biggest fear was, is this thing going to be big brother? This patient-clinician interaction is a very vulnerable one. Is it going to grade how I provide care? What if I mispronounce a med name? Is it going to watch over me? So we've never watched over clinicians, but there's things you can do in the user experience to also win that trust as well.
for our system, are, I won't go into the technical aspect of things, but we retain no data. We have a zero data retention policy. don't watch how the clinician practices. There's no sentiment analysis between patient and provider. They, know, it's literally, you know, sunk and then, you know, wiped away, if you will. But the user experience matters extremely important in order to gain that clinician's trust because
who wants to be washed over while they practice care, you know.
Tom Maxwell (23:35)
By the way, doctors are doing this all over the place. mean, every healthcare facility, mean, the vet clinic, I take my dog to the vet and they're doing it ambient. mean, they're doing it 100 % of the time. And all, this is not about the police, this isn't about scrutinizing your work or any of that. All we're trying to do is make it faster, easier, and better, right? Faster, easier, and better, meaning like, you're finding out this information, why do you have to go type it up now? You found it out, you've talked about it. Verbal, by the way, is way better communication than typing. promise you, you can talk faster than you can type.
I don't care who you are. My mom types 95 words a minute. I guarantee you she talks faster when she types because just keeping up with her is hard. can talk at 200 words a minute. So why not use that as a tool? Why not use it as a tool? That's what I keep always going back to is if we can make it easier, faster, better, and you don't have to go home and do pajama documentation.
Life's way better. That's really cool. Talk about burnout. What have you seen, clinical burnout? mean, this has got to help solve that problem, right? I I hear all the time that home health and hospice is too hard because the documentation's too much. I'm going to quit and go work at a wellness clinic or I'm going to go work for, you do IVs or do something different. So we see nurses leaving our industry. Do you think we're solving that with this?
Alex Milani (24:45)
Definitely addressing a large portion of it. I I think the turnover in home health is maybe the highest in all of post acute or almost all of healthcare. a reason for that and a lot of it's to documentation, probably the number one reason if we had a pinpoint one. This is directly solving that. So there's a variety of
where directly solving then indirectly solving.
Burnout is probably number one or number one, you are spending more time looking at the patient, actually assessing the patient away from your computer. The clinician likes that, the patient likes that. That's a factor in burnout. But in addition, you know, when you get home, you can actually unwind and not document for the next four hours after, you know, you pick up your kids or make dinner or anything. Like you actually have your time back. So.
No, definitely burnout is probably one or two of like our top variables we measure.
Tom Maxwell (25:38)
So Alex, I want to.
After you get this thing rolled out to home care and racist customers over the next few months, I think is what you said. I want to grab a few of those clinicians and bring them into a room and let's bring them onto this podcast and let's talk about that experience. I think it'll be awesome. Clinicians want to hear what other clinicians are doing. So let's make that a plan. We'll grab a few from your first customers and let's talk about the experience. And it's going to get better and better and better. I mean, everything evolves over time and it's going to get better and better, faster and faster. And as it learns more and the clinical intelligence portion.
Alex Milani (25:51)
would love that.
Tom Maxwell (26:08)
It's cool. What else is on your big roadmap? I I don't want you to give me all your secrets out, I know your brain's going crazy. and I were just, Alex and I were just on a conference call a little bit ago. We were talking about some of the other cool AI tools that are out there around recruiting and finding nurses and some of that kind of stuff. It's blowing my mind every day and I spend a lot of time and a lot of tokens looking at data. But what else is on your mind out there?
Alex Milani (26:23)
Yeah.
Yeah.
Yeah, I mean, there's a lot of things I want to say and I'll, I don't want to give away all the secrets of course,
having a more clinical background, my initial reaction is always going into patient care specifically and how can we empower that clinician to provide the best care possible as it relates to risk.
What can we surface to that clinician and say, know, Mrs. Jones has a 30 % chance of, you know, readmitting due to her heart failure. You know, this should be addressed. That's where I always like to lean, but there's other things as it relates
know, more accurate documentation upfront, reducing the need to do this QA and coding economy that I find is very fascinating. It's actually one of the reasons why I wanted to do something so that what maybe stay in post-acute and home health specifically is there's this.
very interesting economy where, I'm a home health clinician and, you know, I just saw this patient for two hours. Now I'm in the driveway for two hours documenting, you know, without AI. That form is then sent abroad usually, or maybe to some other department and they're fact checking my work. I always found that fascinating because on the acute side, that doesn't really exist that much. And then 72 hours later, I get back someone's edits. So being in the...
point with the patient, with the clinician, we can reduce some of that downstream QA encoding that we're very, very passionate about going forward. But those are the things we're thinking about in a macro perspective. There's a lot of blocking and tackling we need to do with this launch. just on a broad perspective, there's a lot of interesting genres or economies that exist outside of this point of care that maybe can be solved. We'll see.
Tom Maxwell (28:13)
I was listening to a podcast this morning on the All-In Podcast at the gym.
And so the Nvidia CEO is on there. You probably know him. He's one of the richest companies in the world or wealthiest companies in world. And they were talking about, AI going to replace jobs? And everybody said, and they predicted that AI was going to replace radiologist. And we didn't need radiologist anymore because AI is way better at reading. And what they found is it's actually the opposite of that because the AI is so good.
The radiologists are interpreting these x-rays and scans so much faster. They're able to do so many more of them and the return result rate is so quick. So you do an x-ray, you do a CT scan, you do an MRI, whatever, within minutes that one, it's reading it, and two, the clinician or the radiologist is validating it, and three, now they're treating that patient quicker. They're not waiting two weeks for it to go from the computer to the system to the radiologist, back to the doctor, back around.
down to the onychologist, they're now taking care of you until you have cancer or whatever. It's so much faster. So now they're treating more patients and so it's the opposite. More radiologists are coming into the space. More radiologists are staying longer. They like their jobs better. It's pretty bizarre. they're just talking about the jobs aren't going away, they're going to just change. And the speed we can do it. More patients can now afford to get a scan. More patients can now get a scan quicker. We can order them and process them a lot faster. And so we're treating more people.
Alex Milani (29:32)
Yeah, yeah, exactly.
Tom Maxwell (29:43)
I mean we're the sickest country in the world, or also the richest country in the world, but we're the sickest country in world, and there's a way to fix this. And so, you start talking to those guys, like the Nvidia CEO who's making all the chips in the world now, I mean, they're really bullish on this industry and what's going to happen, especially as it relates to AI and healthcare. We're a small portion of that. We're a very powerful portion of that. So, it's pretty wild.
Alex Milani (30:04)
Yeah, agree
more. The one thing I would just say is, not agree more that there's probably an overreaction that AI is gonna replace all jobs or whatever, but the only ones that are gonna be replaced are the humans who don't use it. And I think that's very important as it relates to radiology in this example.
⁓ You can be more accurate, that's better for patient care. You can see more charts in a day and also probably reduce some cognitive load because it's helping. So the only jobs that I see being truly replaced in healthcare are the ones that don't use this type of technology.
Tom Maxwell (30:43)
labs, another good area. mean, it's easier to order labs and know what labs you need and make sure they're focused. And I mean, once you have the blood sample, you can do so much more with that now. And so they're talking about the volume of lab tests that are being ordered and being processed and being read appropriately is completely different than it was even a year ago, two years ago. So we're able to diagnose and treat diseases a lot faster, maybe even catch things quicker. Pretty awesome. I'm blown away by it. It was a great.
It's all in podcast is doing a live thing and they had him on there. It was pretty good. He's a very impressive person. All right. Talk about the team at Steno. Tell me more about that team that you hired. I know you got a young, great, entrepreneur, energetic team working hard. Probably some of the hardest working guys I know. Tell me about them.
Alex Milani (31:28)
Yeah, definitely. Man, we have the best extremely, I would say engineering, AI heavy. That's by design. Kind of all across the teams, all across the world, all across the We're actually having an offsite right now, so all of them are in this room over there. But I mean, man.
The engineers that we have are just magicians. I don't know how else to explain it. They're like the Michael Jordans of our team, if you will. Their understanding of artificial intelligence and where it's going
just like, I've never spoken to two smarter So we've been extremely lucky from the hiring pipeline on the product engineering side. But I mean, these guys...
Do not quit. I mean, they're working 24-7 to push releases every single week that I think any other company would take them two months to do. just pretty miraculous at that speed and turnaround that everything's been going these past couple years.
Tom Maxwell (32:27)
Yeah, it's pretty wild when you start looking at it. And when you get a team, it's...
that has the hustle and passion, I mean, they're really solving a world's problem, right? They're going to fix 190,000 people's biggest headache, right? And if you could prescribe one medication for that, it'd be amazing. so, when your team is successful, there's no doubt they're going to be successful because they've already proven that. But when they're successful in dominating this industry with faster, quicker, better documentation, it's going be amazing. And nothing more exciting than being on a team like that when people are really excited about what they're working on.
Alex Milani (32:39)
yeah.
Tom Maxwell (32:59)
could really show results Think about how many lives are going to be, you mom's going to be able to come home and not have to do documentation or dad's going to come home and not have to do documentation for the rest of the night and following up with patients is going to be easier and looking at what's going on with my patients is going to be easier and looking from visit to visit is going to be easier and so yeah, you guys are working on is really cool. Should be very proud of
Alex Milani (33:18)
engineer's job is oftentimes goes unsung. They don't really understand the impact necessarily. We always try to loop them in with customer calls so they actually truly understand who that user is, who that clinician is, what their two pain points are, not just so they understand the clinician more, but also so that we can track things over time and see how we're improving. So we always try to make sure our engineers are actually chatting with users as well.
Tom Maxwell (33:42)
Yeah, what is your thoughts around, and I know your engineers are using all these tools,
There's all these new tools coming out with all these new models and development tools and all that. mean, just keeping up with them all is hard, right? I mean, for me to do it. I I have probably more time than your guys do, but is there a favorites out there or how do they keep up with all the tools that are changing in the, I mean, it's just, it's crazy. It used to be you were a C++ developer or you're a COBOL developer or you're a Delphi developer or you're a SQL developer and now you're a full stack developer. mean, now it's like, holy crap, it's a, you're an AI prompt engineer developer.
Alex Milani (34:17)
Yeah.
Tom Maxwell (34:17)
I
mean, it's so crazy.
Alex Milani (34:20)
Yeah, it's crazy. thing, the amount of things that are getting released, uh, the only thing that there's more scribe, the only thing that there's more releases, uh, in a week is besides scribes is probably new AI tools for engineers. but every Friday we do a deep dive session. It's only 30, 45 minutes with the engineering team. also try to determine what's hype, what's not, what are people using kind share that brain aspect of things as it relates to.
you know, Claude code or some of these newer tools that are coming out and how people are using it. That's kind of how we try to keep up to speed. But the main one that I don't think we're the only ones out there Claude code has been like game changer for engineers as opposed to throughput. There, you know, it's not a coincidence that our product features have been like exponentially increasing starting around Thanksgiving time. That's when it was really released and it's only gone up and up and up and up. So, you know, an engineer now.
is like three engineers in one almost. And you know, I only see that getting better and better, but yeah.
Tom Maxwell (35:21)
Yeah, it's crazy. You used to code your code. I mean, I was a...
you know, back in the days was doing a bunch of that. So you would code the code, you would hand it off to QA, QA would QA it, and they would hand it off to a use case person. Use case person would then call it whatever you want, would go build the test case scenarios for you, you'd go run those test case scenarios, turn the code back in that you wrote two weeks ago, they would give it back to you, go fix all the bugs you had in it, go back to QA, back to use case, back to test scenarios, and it was just a circle, right? And I'm talking about waterfall and all that, and now we have agile, and agile's even slow now.
Nobody's doing agile like they used to. Agile's like, there's way better ways to code than agile. now the same engineer that's coding this stuff using Claude Code can tell it to build the test case scenarios and go QA itself and come back with five improvements. And then, back in the day, Matt Casero used to work for me. One of his big jobs when LHC and our big customers was he was doing query plans to make SQL queries faster. We don't need a Matt for query.
Alex Milani (35:55)
Yeah.
Exactly.
Tom Maxwell (36:25)
I mean, the code will do it itself. You can just tell it, make this into a better query plan, make it go faster, use less tokens. It's absolutely amazing what we can do now with coding.
Alex Milani (36:35)
it blows my mind, like every week something new, it is just insane the amount of throughput and time savings and efficiency and just more in every single aspect as it to the engineering team. It's miraculous.
Tom Maxwell (36:48)
Yeah. And then I think product scoping and, and you and I were talking about the lovable app or the Claude code app is another one out there. Figma. to, to get a working prototype, you can do that in hours. What used to take us weeks and weeks. And we had these UI UX engineers that would, you know, come in and make everything beautiful in colors. Now they just, you know, they're writing a quick prompt saying, Hey, here's the color palette of what we want to use and go make all these buttons and make this button move over here and go over there. And I mean, it's, it's literally done with just prompts. Now it's not, you don't even have to,
Alex Milani (36:53)
Yeah.
Tom Maxwell (37:17)
do it with
writing the code, which is bizarre.
Alex Milani (37:20)
yeah, product
scoping implementation as relates to PRDs and everything has just, you ⁓ has accelerated so much by giving that PM more power to actually build their tool before it actually is built. So when you hand it off to engineer, it's clear there's less questions and they're ready to rock. had explored your four apps this morning. It's like that, that's exactly correct. It's pretty amazing how.
Tom Maxwell (37:47)
Wow.
Alex Milani (37:48)
fast like a PM can move as well now too.
Tom Maxwell (37:51)
All right, let's talk about implementation a little bit. Once you have this thing embedded inside a home care home base, we're going to go out and train 190,000 clinicians on how to use ambient listening in the home to ease up clinical documentation. Thoughts? Is it two hours, four hours, 10 minutes? How long is it going to take? How are you going to do it? What are your thoughts around that?
Alex Milani (38:12)
Yeah. the implementation on boarding is easily the most important part of all I hate to go on like a little spiel, but, home health is extremely unique, uh, as it relates to healthcare. When you're onboarding, maybe a scribe, we'll just pick on scribes right now in the acute side. say I didn't pay attention during onboarding or whatever, I can easily just tap my coworker Susie next to me like, Hey Susie, do you mind?
telling me how to do this, but in home health, they're on an island. Oftentimes there's no wifi, but you know, in the acute side, IT's down the hall, you're connected to this beautiful wifi. So onboarding and implementation is the true unlock to do this at scale. So we do not take that lightly. We fly on sites, do days and days and days of onboarding. We have 24 seven support within our system.
And at the same time, you know, also always offer one-on-one sessions for folks who, you know, wanted another refresher. a big thing we're also doing is adding tutorial mode. This is a new launch coming out soon, which is essentially you swipe it on and it basically is this onboarding through agents, if you will. So it's reminding, guiding clinician through almost like a video game does when you first turn on a video game. It's like, Hey, this is how you jump. This is how you punch or whatever.
It's doing that same thing for that clinician. So when they're in this unique environment in the middle of Texas or whatever, they also get that same level of care. If, we're not able to call them during this time. onboarding implementation is the most important thing as relates to, you know, getting this done at scale properly.
Tom Maxwell (40:54)
That's awesome. Most people underestimate the difficulty of a mobile workforce. know, Nana's got to get a, or Nana's a patient, Betty's the nurse. She's got to come into the office. Instead of going and seeing her patient, she's got to come in for training. She's got to, you know, she's mobile. She's got to, you know.
She's on the road most of the time. She's not in an office most of the time. She can't come in for orientation all the time. And by the way, we still have to go see all these patients. so making training quick, easy, fast, better, self-learning, all those kind of things is going be really important. But also making sure they understand how to use it. And it's going to take some time. We're going to educate patients on it too. Not all patients are going to want that. And so I love the fact that you're doing the consent upfront and making sure that we...
give guidance to clinicians around how to get patients to consent for this and explain what we're doing and why we're doing it and those kind of things is going to be really important. And so I know I've talked to a few other the other CEOs and CEOs that are using your tools and tools like yours and that's an area that they have to spend time on training their nurses, therapists and all that. Let's talk about. ⁓
service codes. So you're using this tool for what service codes? You said home health, hospice, hope, oasis, all that. What about therapy, speech, occupational, home health aids, all those?
Alex Milani (42:07)
Yep.
Yeah. So a big, um, one big difference here between us and some other folks is, um, every single discipline. you know, RNPT, whoever, every single form type and then every single question type. so I, I think I started this podcast out by saying, you know, Hey, I'm going to use the word Oasis a lot. It's kind of a misnomer because only 60 % of,
Oasis document is Oasis questions, right? There's a 40 % that are state mandated, agency specific questions. So we have AI mapping to every single question type, every single form type, every single discipline. So during that onboarding process, folks basically give us all the forms they want to do. We map them, you know, in a couple hours and then we begin the pilot. they can really, you know, do any form, any type of document, whether there's a narrative note section at the end, we do that as well.
basically flexibility was extremely important in the way we designed our architecture because no agency practices is the same. They all have kind of a different flavor and that's kind of how clinical practice is. So yeah, every form type, every discipline, all the questions as well.
Tom Maxwell (43:15)
Well, that's awesome. And I'm sorry that was a lot of work, I know, because we built all those questions and answers and forums. There's a lot of them there.
Alex Milani (43:21)
No, It was a fun project
and it makes us different than the rest, so that's always good.
Tom Maxwell (43:29)
Yeah, Scott Burt and I back in the day in 2005, 6, 7 with Stephanie Chaney and the whole team were building all these forums and questions and answers and.
the compliance side of it and make sure you check this box and do this and validate this. I mean, it was years and years and years of work that's sitting there and I know you guys had to go map all that. So that's fun ⁓ and hard, by the way. really awesome. I didn't know you were doing all the disciplines, so that's great. And physical therapists are different than occupational therapists, which are different than social workers and chaplains and aides. so making it easier for everybody is going to be really important.
had a leader one time tell me that, and this always blew my mind, I'm not going to tell you her name, but she's like, we can't trust our aides with a device. Like, you can't give our aides a device. We need some kind of other technology for them. And I'm like, okay, you don't trust them with a $200 Android tablet, but you trust them to go into Nana's home and take care of Nana? Like, I don't understand that. So we got over that pretty quickly, but everybody needs to be part of that care circle. And you the other thing of this is,
Alex Milani (44:21)
Yeah, that's... Yeah.
Tom Maxwell (44:28)
By doing it this way with the ambient listening and as you guys are doing more things, advancing it, being able to see what the therapist did and being able to, for the nurse to be able to get that carried forward so I could understand what the PTOT and ST did on their visits and they can understand what the nurse did on their visits that Nana has a wound on her right hip, she's got a, here's how we're treating that wound, all those kind of things. And so that co-collaboration that can happen through the ambient listening, through the, as you called it, I think it's the prep,
Alex Milani (44:38)
Exactly.
Tom Maxwell (44:58)
prep for the visit as I'm driving from one area to another and I'm trying to read what's going on with my patient. By having the AI essentially produce that for you and give that to a clinician, it's going to make their life a lot easier and it's going to make that care coordination and collaboration a lot easier as well. So that's super exciting.
Alex Milani (45:15)
Exactly.
Yeah, it's better for patient care too, know, going back to it, you know, having that next clinician prepped and ready to go on, this foot ulcer and what it looks like today, that's only going to be better for the patient too.
Tom Maxwell (45:29)
You know, one of the things that COVID brought us was, and we've evaluated this a lot, it used to be the care continued. We had the same nurse taking care of the same patient over and over and over. And now it seems to be that with COVID, we just needed a nurse. And so you might be the starter care nurse or you might be someone that's never seen this patient before. And so how do I get up to speed on visit eight? Because, you know, nurses don't work seven days a week, 24 hours a day. So if I'm going to go see your patient on visit eight and I've never seen them before, how do I get prepped for that patient?
How do I get familiar with that patient? Do have to walk in and just ask the patient what's going on with them? Or do I get all the details up front, which I think would be really cool as well. And it's not the easiest. I PointCare does a good job of having some summaries, but it's not the easiest to go navigate through that while you're across the So you're spending 20 minutes in the parking lot before you go into the patient's home. And nurses don't have time for that. So I think this will help us a ton there.
Alex Milani (46:22)
Yeah.
100 % and
also one beautiful thing about healthcare is or at least you know our Role in it is we're also patients too, and I you know having been a patient You know it's always annoying when you have to tell the same story over and over and over and they're like isn't that in the chart right? So you know empowering that clinician to you know have that prep beforehand also helps you know with that patient Provider relationship and only helps with star ratings. There's a lot of things low-hanging fruit that are you know
quick wins that are easy solved from an AI perspective or a technology perspective.
Tom Maxwell (46:56)
Yeah, I'll give you a perfect example that happened to me yesterday. So everybody knows that I'm a peptide freak and an IV freak and a wellness yesterday I was at the peak downstairs from my office getting an IV and they were giving me my TB 500 and my BP 157, which are two peptides that I take to help you with recovery and healing. And I was getting an IV and they gave me the forms and they were like, Hey, we need you to fill these forms out. And I'm like, okay, you already know my date of birth. You already know my name. You already know my
Alex Milani (47:24)
Yeah.
Tom Maxwell (47:25)
address, you already know my email address, and I'm just signing a release form saying that I understand I'm taking a peptide and getting an IV, but I have to fill all that crap out again? Like seriously, it's already in your system, like why can't you print the form with it in there like, no we need you to fill it out. So I spent 25 minutes writing about the same thing on the top of every one of these forms so I could sign off on it I'm like come on guys, this is terrible. Like, I know, it's just like terrible. And I know you know this stuff, like by the way.
Alex Milani (47:45)
Yeah, you should get a stamp next time and just stamp it, you know, yeah.
Tom Maxwell (47:52)
You have my phone number, you have all that stuff in your database, just print the form with that on there. Like, no, you need to fill it out. It's a blank form. Like, this is dumb. Like, come on, guys. Yeah, yeah, yeah, yeah. I've been there every week for the last two years. Like, what's changed? Come on. But yeah, it's just, yeah, you're right. Being a patient and dealing with that's interesting. All right. This has been awesome, man. I'm so excited for you
Alex Milani (47:57)
Yeah. Yeah, yeah, there were three days ago and nothing's changed, yeah.
Yeah.
Tom Maxwell (48:15)
Excited to see that you got the Home Care home base contract completed and you're moving forward with this. I wish you'd give me the exact date. You're going to roll it out, but I know that's kind of a little bit of a secret. There's things that are going on. Soon is what we're going to say. know we're going to change the industry. already had some great pilots. People love what you're doing. They called me and told me, like, Tom, you need to figure out how to get involved in this company. It's really good. I know Maxwell TEC and Steno are going to spend a lot
of
time trying to work on documentation and make it look great. Super excited about where you're headed. As a customer, and you own a few home health and hospice agencies and on the board of them, super excited to get this thing rolled out to our clinicians. It's the thing that people really need right now. And I just hope we can solve burnout. I hope we can solve recruiting. I can't imagine being the agency that doesn't have a tool that reduces documentation by 50 % of the time and then trying to recruit or trying to hire somebody that didn't have that tool.
It's going to be easier, it's going to make it quicker, faster, for me to recruit and that's what we want. So congratulations to you and the team.
Alex Milani (49:12)
Yeah, thank you so much. Yeah, no,
signing with Home Care and BASE and having, you know, the direct, know, alignment on where we want to go together has just been amazing. And we're so excited to be launching this in a few, you know, short months.
Tom Maxwell (49:25)
That's awesome. So congratulations. Alex, any questions you want to ask me?
Alex Milani (49:30)
No, I mean, I'm just excited to be on this Well, here's a question. When do you actually start the podcast? Like, why do you start the podcast?
Tom Maxwell (49:39)
I went through this a little bit on one of them with Kristin. I was in North Carolina and I sat down with the CEO of AuthoraCare Kristin. She started like...
I'm just like yelling at everybody. She was scolding me like, you need to do this, Tom. You're the reason you need to do it. So went home and talked to Jennifer about it. And Jennifer's like, you should do that. You're absolutely the right guy. know everybody in the industry, everybody likes to hang out with you. All you do is talk. And it's really all I do all day long is talk. And she's like, you like to interview people. You like to talk to people. And you should do it. And I'm like, I don't want to do it. So I said, OK, Jenn, the only way we're going to do this is if I hire a real producer. We get all the right stuff. I don't want to have to do the work.
Alex Milani (49:56)
Yeah.
Tom Maxwell (50:22)
wanted to schedule the meetings and get all that. And so we hired Cole and Cole's been absolutely fantastic. He's way exceeded my expectations. mean, he used every AI tool in the world to build a website, produce the shows, do the clips. He used Riverside to produce these podcasts. It works great. Podcast quality's good. We have over 100,000 impressions. People love what we're doing. The feedback we're getting is awesome. And so here we are, we're doing it. The last thing I want to be is Joe Rogan of Home Health.
It's been a lot of fun and people are starting to learn a lot more about our industry just by sharing amongst each other on the podcast. So we have a bunch of really experienced guests coming on, but I like to get them from all different angles, right? I want nurses, I want chaplains, I want social workers. I had Dr. Mayo on, Dr. Richter, who are medical.
directors for hospice. We do some hospice, do some home health. Gonna do some palliative care stuff here coming up pretty quick too. So excited about that. So yeah, it's been a great one. Yeah, yeah, guy's a cool guy. And his story about his wife and his current wife, Cindy, is really amazing. So yeah, I'm a huge Dr. Mayo fan. He's one of my...
Alex Milani (51:15)
Yeah. Right, I listened to the Dr. Mayer one. He's a genius. He's truly a genius.
Tom Maxwell (51:31)
biggest mentors and someone that I respect the most. He's a great guy. Some of the stuff they're doing at St. Croix is pretty wild. So, I mean, they've grown like mad and have done just amazing things for patients. Well, cool. Alex, thank you so much for your time today. Appreciate you being here. Again, congrats on your success. It's going to be a lot of fun.
Alex Milani (51:46)
Yeah.
But yeah, thank you so much for having me, truly honored.