AI-based Smart Documenting with JotPsych

Jackson: We see our mission here really as we wanna take this technology and make it so that it can help you as the physician, as the clinician. get all of the stuff done so that you can focus your own mind and your own training and your own background on doing the best, caregiving and diagnosing possible.

Bruce Bassi: Welcome to the Future of Psychiatry podcast, where we explore novel technology and new innovations in mental health. I’m your host, Dr. Bassi, an addiction physician and biomedical engineer. Today we are with Nathan Parum and Jackson Befelt from Jot Psych.

Nate is the CEO and Jackson’s the CTO, and they can tell you . What those mean and what the roles are. Jot Psych listens to your client interactions and generates complete HIPAA friendly, medically accurate notes, which we’ll talk about later, which is kind of astounding and like magic in my mind. it records your patient interaction on the web or in a mobile device.

In person or via telehealth, and then you can review and edit the auto-generated note and the template that it goes into, and you can basically copy that into your EHR system. So welcome to both of you.

Nate: Bruce, we’re 

gonna pull you into our sales team. That was a good pitch.

Bruce Bassi: Thank you so much. So documenting in my mind, is one of those things that gets a very little attention, but it’s probably takes up a disproportionate amount of your time and energy and frustration. there’s no class on documenting and you just learn afterwards by trial and error and anecdotes and when things go wrong.

Why do you feel that . Life is this way for us, clinicians like why is it so important? But yet we spend no time on it.

Nate: Yeah, it’s a really interesting question and. In order to understand how charting got to be the way it is, you actually have to go back 10, 15, 20 years to when the US government was getting really excited about electronic health records. And, know, at that time government was saying, Hey, if you want to get reimbursement, full We actually require that you be on an electronic health record. And so what happened was there was this explosion of software companies that were trying to make, healthcare records that, could be easily, you know, interoperable. So records could be sent across health systems and across clinics, also, uh, comply with all kinds of, know, both private and public payer mandates. And probably started out as a really good idea, morphed into this monster, that, you know, has in varying degrees, complicated and benefited, patient care. And so, you know, it’s one of those double-edged swords where, one hand we’ve got a lot of great data from EHRs. but on the other hand, nearly every clinician we’ve ever spoken with misses the days of paper charts if they’re old enough to remember those.

Bruce Bassi: Yeah, it’s not kind of a weird thing. 

Jackson: No, I was just gonna say, it’s just one of these situations where, you know, the burden disproportionately fell on clinicians to take on all of this extra documentation and what was kind of promised as, oh, you know, it’s digital now. It’s gonna be easier. We’re gonna be able to. You know, move things around more quickly. I think most of the people listening to this probably know that interoperability promise has definitely not come true. And what’s really happened is the amount of charting that needs to be done for every single interaction has just ballooned the

last, uh, know, several decades. we have, you know, people telling us that They, see a full slate of patients every day and they’re trying in between those patients while they’re like eating a sandwich to like type down some stuff. They’re typing or writing down things while they’re in the room with the patient or while they’re on a call with the patient. And then still after all of that, they’re going home and they’re spending an hour or two after they’ve made dinner and are trying to spend time with their family. Keeping up with their notes, and if they don’t, then they’re chronically behind and you know, everybody in the field knows somebody who’s chronically behind on their notes and they know just like, what an absolute about a mental health issue. What a problem that can be for the clinicians as well.

Bruce Bassi: Yeah.

Nate: Yeah,

Bruce Bassi: going through, my experience as a clinician in a large hospital system, I kept encountering these scenarios where . there’s a team or some sort of administrator who thinks it’s a good idea to add this, like new screener or this new element to the patient chart. And over the course of a few years, the note just like balloons up and even larger and larger every passing year with a new, type of thing that you have to include in your note.

So, I think one of the things that people . When people say they miss paper charts is that they can’t be told like of these, like new elements that they need to include in those paper charts restricted to basically the core elements of what they think is clinically pertinent, but then insurance companies and improvement, They add these other parts to the note that are important to them and as a result, the clinician needs to now make a note that’s . as an audience of maybe six different teams, like the insurance company, the biller, quality improvement, the patient, et cetera, you know, other clinicians that they might be speaking to on the other teams, do you have any data as to, this is such a big burden for clinicians and why you felt it was necessary to address.

Nate: Yeah, we think about this problem as the, kind of eternal trade off between structured and unstructured data. So. If you think about a spectrum, all EHRs, all documentation systems, even outside the medical field, follow along. one end being like a blank word document where you could just free flow information. then the other being a really structured survey where you’re, you clicking, predefined, Preset data or, inputting and EHRs vary in their degree of structuredness, but a lot of the note bloat that you’re talking about actually comes from lot of administrations. Saying, Hey, we want more structured data. And so you get these like really, really interesting evolutions where, know, there’s an EHR out there, we won’t say which one. That was actually like a golf course management software that then transitioned into medical software because setting up a tee time and like putting in all the metadata for like starting some time on a golf course was actually similar enough, methodologically, to the way that medical charting had become in order for that transition to make sense for them. and so one of the things that we find is magic about this point in history. Is that artificial intelligence now allows us to take unstructured data and turn it into structured data without involving a human being. That is like the key thing that ai, like if you had to say like, what is the key thing that AI is allowing for?

It’s, it’s, and I grew up in a medical family where I saw this burden. You know, my dad’s a neuro oncologist, my mom’s a nurse practitioner. were doing a huge amount of medical charting. when I was a kid and it affected them and it affected our family. and when we started looking at kind of the available tools and also the impact that charting was having on things like, actually performance of medical care or, you know, clinician burnout, it was a problem that we really felt to tackle and, called to address.

Bruce Bassi: We should give the listeners some time to Google Search Golf course EHR system so they can figure that out. 

Jackson: It has a

very obvious name. we don’t think they changed the name in between, 

Bruce Bassi: there’s gonna be a spike now in traffic in that. Go ahead Jax.

Jackson: No, I was just gonna say, I mean, everyone listening to this knows that charting’s a problem, and most people even say it is probably the biggest problem with regard to burnout. And this is across, you know, all specialties in the medical field because when someone goes into medicine and patient care. They are not thinking, oh, I’m gonna just be doing a lot of like clerical documentation work. You know, you go in to actually have FaceTime with other real human beings that you want to give care to and you want to help. And then what ends up happening is that in order to do that, yes there is a component that is actually, you know, medically relevant and is important to take down notes and make sure that you remember things and document that. But I think everyone has a sense that a lot of the stuff that’s taken down Is just kind of that bloat, it’s necessary in some way, in some systemic way, but it takes away so much time that people are starting to wonder like, okay, is that minor, benefit that comes from me filling out 15, 20 minutes of paperwork for every single half hour session that I’m doing actually pushing the needle there?

Or is it just making my life so crazy that it’s then hard for me to like care to people? And I always come back to the thing about know, when I go to a doctor, even like a primary care physician, I do not like that I’m in the room with them and they’re not looking at me, I’m talking, they’re asking me questions, I’m talking to them, and they’re just totally turned to the side and typing, the computer.

And I understand why they have to do it. I mean, if they didn’t do that, they’d never get their job done. But from the patient perspective, It’s so frustrating because that care that you’re looking for, both in terms of the medical care but also the like, emotional care that you’re getting from a provider, that just feels like it.

You know, it’s not there sometimes, and that’s one of the things that really proud about, being to bring back the clinic room with our software.

Nate: let me just put some numbers around that. ’cause a lot of people. When they think about their relationship to charting, they worry it’s actually a personal problem or they worry that it’s a personal failing or they worry that their perspective is unique. just a few numbers to kind of frame this.

So the thing that Jackson’s talking about, which is patients don’t like when you’re on your computer. a, an encounter, they would prefer that you focus on them. know, about 25% of patients feel like they’re not being heard. you’re on your computer or you’re jotting something down when you’re in front of them, physicians, about third of them feel that, they’re doing just way too much charting or they’re, you know, they’re charting just an absolutely excessive amount. this is the, this statistic that gets me 34% of all medication errors in malpractice suits are related to typos in an EHR. Lemme just say that again. A third of all medication errors that get litigated or settled in a, judicial way come from human error in EHRs. And I think that for us, what we believe is, just that people don’t like it, it’s not just that it’s bad for patients. it’s actually like not good are generally not, great at, down accurate information after 15 visits, in a day. So, there’s just so many good reasons fix this problem.

Bruce Bassi: it’s a huge problem I think a lot of people can relate to this, in disclosure to the audience, I’ve given a, a shot, I’ve tried Jot Psych and I think way it works and picks up on, even the small details. Hopefully that can, take me out of that 34% if possible, down the road.

Knock on wood, it’s like magic. it really does pick up on all of details that were, discussed in the session and puts it in an organized. methodological fashion depending on whatever type of template that you, want it to be in. And I think that’s probably one of the challenges that you all had, and you can speak to this maybe, is when think of they, they it as a somewhat personal, in the respect that have their own style, they have their own, structure that they’re familiar with, that.

they’ve probably honed in on over the past 15 years or however long they were practicing, and they’re wondering, how can a computer start to learn my style? Is what, is the trade off here going to be? Because yes, I agree with you that a time sink, but. What would life be like with this AI system kind of incorporated?

Can you give a audience member, kind of a visual of how, what that would look like and why it would be for them? 

Jackson: Yeah, I’ll jump in here and say that when you think about the evolution of ai, and I’m gonna start with, you know, machine learning. This is kind of what we were calling this like same pattern of computation a couple years ago. The first thing that machine learning actually proved to be really good at, as far as like general mass consumer facing software was, was these recommendation engines. You know, when you think about how YouTube changed from 2008, when it was just a place where you hosted videos to, 2016 when people would just load YouTube and be shown things that was relevant to them,

that was a huge victory for machine learning because before that curation process, that ability to kind of. Learn what you like and then deliver something to you. That was something that humans had to do and you just couldn’t do that for millions of YouTube users. look at all the videos they watched and then pick out a couple and give it to ’em. That’s what your friends did, a lot of years. You know, that’s what we do socially, but like machines weren’t good at that until we kind of crossed this technical Rubicon. Into something where like the machine can take in all this data and then kind of filter it and figure out like, okay, this is what you want. And then that is really one of the big wins in AI now, and especially with these large language models, is that this is something that it’s very good at.

As Nate put out before, like taking unstructured data and turning it into structured data is a huge strength. But also in a whole bunch of little tiny preferences. Oh, I like it this way. I like it that way. And then being able to kind of like learn the patterns from that and apply those to just an individual practice or an individual provider. And that’s something that we’re really focused on with our software, is understanding that every single provider, you know, regardless of their specialty, they have a certain way that they like to do their notes. this is something, people come to us and they say, it is great. I like the idea that I could have a medical scribe that’s always in my pocket and always ready just listening to me and generating these notes.

But, I like my notes the way I do them. I don’t really want to just like have to adopt like whatever, structure is coming down from on high. And we think of that as every provider having their own, like personal fingerprint. and we try to build our software in such a way that, it lets you take your fingerprint and get the notes to be generated in the way that you like it.

Whether that’s, you bullet pointed information or you want a long free flowing narrative of information or. You only want just like the three most important things with just like quotes directly from what the patient said, or you want 16 paragraphs saying every single tiny thing that happened, whatever your preference is there. We build the software to try to like hone in on that customization so that you can feel like, Hey, you know, this is like a human medical scribe. They’re just way faster, way more accurate, and about one 30th the cost. that’s what we’re trying to build, and that’s like the most exciting thing as far as, you know, our product goes technically is that customization aspect.

Bruce Bassi: is that customization aspect technologically more easy or the less challenging piece to this? Or are there other, challenges, in this process developing a, outcome for the clinician?

Jackson: Yeah, I would say the customization’s definitely one of the biggest technical hurdles, and it’s something that, you know, we’re constantly making progress. On, as we’re recording this, we just released a new feature that allows people to build templates, you know, a lot more carefully to the kind of sections that they want because we know that, some people they just really need like an HPI section, but some people, you know, they need the HPI here and then they need the topics discussed here, but then they want this here and that there, and, we’re giving the ability to kind of like Lego, like, build those together. But the next thing that we’re working on. Is this really exciting feature where you’re gonna be able to just talk to our software and say in kind of a vague way, like This is the kind of note that I need, or I need a section that does this kind of thing. I don’t really know like exactly how to like train it, but like you can have a conversation with our software and our software will help you build your own little Lego blocks that then you can then use in the future to generate all of your notes. S.

Bruce Bassi: In terms of . Uptake and adoption with this model, do you feel? Is it the clinicians, what is the, determining step here? what is the, bottleneck, in terms of why more people aren’t adapting it now?

Nate: I love that question. we think about our whole business as a rate limiting function or rate limited function. Right now the biggest rate limit is just people who know and feel safe using ai. AI has a brand problem, which is that there is a portion of the population they themselves, a certain fear or trepidation about using AI tools or they perceive that their patients might. of that has to do with privacy. Some of it has to do with, just the black box that is ai. And so, you know, I think the biggest thing that we’re trying to do folks is introduce them to the tool, have them use it for free, and walk them through in as rich detail as possible, how we protect things like patient information, security of that. once people start using the software, we can see this like if someone creates more than like 15 or 20 notes, at least in our software, they generally stay using it, indefinitely. mean obviously we’re a young company. we’re not even a full year old. but we’ve had, many, many users who tell us that this has changed our life and they continue to, of increase their usage of the software over time. So awareness is, I mean, to answer your question, pointedly awareness is the problem.

Jackson: And one of the things we often hear from our providers is that they themselves are sometimes shocked at how willing their patients are to have a recording going on. especially when we started working With behavioral health specialties, we thought, oh, there’s no way that this specialty is going to, you know, at a cultural level be okay with like a recording device being in the room.

And, you know, sometimes it’s not, you know, there are certain patient populations where that’s just, uh, a non-starter. But what we found is that maybe 2% of any of the behavioral health specialists that we’ve ever approached here, because what patients realize is that you can say, all right, I’m using this thing on the table that, you know, it doesn’t send your name or any identifying information about you to anywhere. It’s just writing this stuff down and then helping me do my charting so that I can sit here with my hands, you know, folded and my eye contact on you, and not have to be typing things or writing stuff down, or otherwise distracted about getting through my notes. So, you know, the way that people communicate that to their patients is, you know, different in different ways.

But like we have been blown away by the amount of patient support that has existed around the software. You know, patients are just as excited about it as providers sometimes because they say, great, now you can focus on me and, you don’t have to worry about what you’re typing over there.

Nate: if I can just make the, just a bit of startup lore here within the company. So when we started We told ourselves, Hey, we’re gonna, we can go after any specialty in medicine, but not behavioral health. we’re actively not going to market to behavioral health folks. We’re not gonna do testing with them. there’s just the information’s too sensitive. So, there was a point over this past summer where we had probably 50 people actively using the software for free. And we said to ourselves, okay, let’s We need to start charging for this because we’re spending a huge amount of our money on, uh, you know, servicing the software and maintaining the APIs and know, and obviously like Jackson and I are working full-time on this as well as another developer named Jacob. we decided, okay, we’re gonna actually go to users and say, Hey, we’d like to, charge. The only people that stayed were in behavioral health. So in some ways it’s just, we got really lucky we worked with, some early psychiatrists, in our, you know, in our journey. and they were gracious enough to support us, when we were still developing the product. other thing I’ll just add as, a misconception, so one of the misconceptions, which is know, what we just talked about, is patient comfort with the technology. There’s a second misconception, is that People hear about our software and because everyone knows Dragon, they think, oh, it’s speech to tech software. I know that I’ve tried it. I don’t like it. And it’s, taken a lot of communication to help folks understand No, no, no, no, no. You can do dictation with our software. We call it like dragon on steroids. you don’t have to, it’s not speech to text. Exactly. You can monologue organically and you’ll get a note. But it’s taken a lot help people understand, no, this is actually a full scribe. It’s like a human scribe, but better in terms of accuracy. and it’s priced at cheaper than Dragon.

Like that is just really confusing to people who have known Dragon in the ecosystem for years and years.

Bruce Bassi: It’s interesting you mentioned, how the health specialist had been the ones who stayed, or for the most part. is there anything else regarding how Jot Psych works and operates that lends itself well to behavioral health specialists terms of vocabulary or how it picks up on other topics that are communicated in a session?

Nate: Yeah, the biggest. Advantage of working with behavioral health, as a specialty, is that the narrative, like the patient’s conception of their own story is material to their medical care. and the majority of the node is actually constructed based on the conversation. really different than, you know, I’ll give you an example.

We were working with an oncologist, up in Boston for a little while and, he was using the software every day he was telling us how much he liked it. then we started to ask him like, Hey, how much of the note do you actually, get from the conversation? And he said to us, oh, like, I don’t know, three, 4%.

Because most of the note that I write as an oncologist is a reflection on the forensic data I have from labs, tests, pathology reports, scans. so. Your technology is fine, but it’s actually not writing most of my note. And we wanna be in the business of writing most of the note because that’s where we provide the most value. what we believe is that certain fields, behavioral health as one of them, palliative care, we actually think is another. that gynecology actually also might be another field where this is particularly important. we wanna stay really focused specialties where the actual conversation is really medically relevant.

Jackson: if I can put this in like a bit of a nerdy, like informational theory kind of way, I’m the technical brain here. When I get excited about this software, I think of it in terms of, when you have a specialty where so much of the information that’s medically relevant is actually happening in conversation. grinds my gears to imagine all that information being transmitted in this conversation and existing in the room, and then having to then be later, like re-written down, just like reconstructed through thought. think of our software sometimes as just this little box that sits in the room that just, it captures all of that information that’s coming out in the conversation and just makes it so that you don’t have to redundantly. Say it again.

If you’ve already said something in the conversation and something comes out in that two way with the patient or with the client, that then exists, it is captured and then is put into the medical note so that you don’t have to have this kind of like crazy redundancy of information theory.

Bruce Bassi: Yeah. If you want to think about the most inefficient way to do something, it’s like instead of having a court reporter, how about just have them rediscuss it afterwards, later on to talk about what just happened.

Jackson: Exactly.

Nate: Yeah. Yeah.

​

Bruce Bassi: what about, so in terms of the oncology example, and I think some psychiatrists might be wondering, okay, well what if I have an inference that I made, according to the conversation, or I want to give my own opinion or assessment,

how would I, um, add that in and incorporate that into the Jot psych workflow?

Nate: there’s three ways that I would say, so The first way is kind of the hacky way that we’ve seen, providers do it. So we worked with a pediatrician out in, City and he had this practice that we thought was really cool where he would go into the room, have an interaction with, you child and the, the would walk out of the room and we just continue dictating into the recording. And what that did was it allowed the audio that we were synthesizing and, and studying to be both the full conversation as well as his kind of private reflections. so that’s the first way. the second way is you can edit your note dynamically in our software. So, you know, Typically there’s only like two or three things, know, that we find people adding afterwards, they don’t wanna speak them out loud. and at a click of a button you can make the change, click save, and now you’ve got a note that the spinal cord we generated. The full note. and the full flesh and body of that note are, you know, partly contributed by you. And we always are, we’re really upfront with people.

Hey, our software does not eliminate charting for you. It reduces your time charting by 90 to 95%, we’re not promising that you never have to touch your chart. the third thing is a new feature that, Jackson can talk more about, but, Over the next, you know, six months or so,

we’re probably gonna be rolling out, a tool that allows you during the visit to jot down like very small things, and then click submit and you submit that kind of private, those private thoughts that you might otherwise type on, like a private notepad, um, into engine, which then we will kind of weave that together with the content from the client visit.

or the patient visit. so we’ve got a few different ways you can do it, but We always say this to people, it can’t read your mind. we have not created a mind reader yet. If we did, that’s probably the business we’d be in.

Bruce Bassi: it also sounds like there’s a process in which algorithm learns your preferences over time and improves. is that right? Based on what you were saying earlier, Jackson.

Jackson: So we have a couple different ways that works. Some of it is actually just giving physicians and providers the ability to go in and kind of, modify what they want. You know, create their own template, set stuff up ahead of time. And you know, we find that a lot of people. In this space, they like to manage things.

we’ve seen people with, dot phrase templates in Epic or something where you sometimes wonder if they would’ve had a career as a data scientist, you know, and been more happy because they’ve created this like, massive infrastructure and they’re like, oh, I made this for my department and we shared it with this other department.

And,

you know, we know That those

personalities exist too, like . to get in there and customize it. So we like to give people all the tools they need in order to go ahead and do that. But then what we also have is kind of a simple system where after a note is generated, on every single section of that note, there’s a thumbs up and a thumbs down. And if you like the way that it came out and you’re happy with it, you can give it a thumbs up. And then our software kind of incorporates that. if you don’t like it, you give it a thumbs down and then we know, okay, there’s a problem here. And the next step, after that, and this is the future we’re currently developing, is being able to, when you give that thumbs down, have a chat box come up and say, okay, what didn’t you like about this? Then you can actually kind of have a two-way with our software there and say, this was pretty good, but it missed this one thing that I think is very important when I see, you know, pediatric patients. And then it says, oh, okay, let me, let me redo this. gonna regenerate the note.

Okay, now here’s the new version of it. Did I get this right? And then if you’re happy with it, then it says Great. I’m gonna remember that for next time. And now you know you’ve improved your own personal scribe with some, directive feedback that you’re giving and in a conversational way.

You don’t have to go in and like open a page with 30 buttons and check boxes and switches and all of that. In order to get this very specific outcome, you can kind of just talk to it like you would an actual human scribe, give them some feedback and then they’re gonna incorporate it.

Bruce Bassi: I wanna go back to your stories real quick. I think out of excitement for the technology, we jumped right in, but a psychiatrist, I like to kind of understand where people are coming from and I think there’s a little bit to be learned here for any individuals who want to into, innovating and, and mental health in particular, When you’re thinking about this idea, or just if you rewind just prior to that, ’cause documenting really isn’t the shiny, glamorous component to mental health or healthcare in general. How did you see that there’s an opportunity here? I think we, Rewind back prior to the release of chat, GPT.

I think the concept of having a technology that sits in a doctor’s room and is able to transcribe the note would just sound completely sci-fi to most people. you took that

idea and, said, no, we’re gonna do this and. Then ran with it and now you’re continuing to improve upon it. what attracted you to using AI in documenting versus diagnosing or treating or some other component of healthcare?

Nate: part of it was that

even though it’s not a glamorous, problem, it is a problem that everyone talks about in medicine. I kind of mentioned my mom and dad are in medicine. Jackson and I grew up in Cleveland, you know, in the shadow of the Cleveland Clinic, all of our friends and family are, in some way by one degree or two degrees connected to the Cleveland Clinic or university hospitals or Metro Health. so I think because of our background, we actually just were aware, we thought of this as a shiny problem. and we also thought about it as a place where we could really make a difference. Given that, you know, Jackson and I are not medical people. I wouldn’t really make a lot of sense for us to go and try and compete in the diagnostic space, for example. because, we’re not, we didn’t go to medical school. and that’s kind of a different world. there are also some really, really cool companies already working on that. So, felt like there was a need. when and I have been really good friends for a very long time, and, we decided we wanted to embark on a project that, employed AI for Good, that, you know, used artificial intelligence to make a positive impact on the world. actually came up with this list and, um, one concept on the list and for each concept, did a bunch of research, we interviewed people, and when we started interviewing people about charting. We very easily were able to develop research partnerships. So the first, like three months of our time on this, company. Jackson and I just lived in doctor’s offices. I mean, we, we just traveled around the country and anyone who would take us in, we would just huddle in their office and like watch them create a note and then, you know, we’d go back to our Airbnb at night and kind of stitch together another prototype. And, was, it golden fun time. and the openness that medical offices had with us. really sent us this message, which is, Hey, you’re onto something that people are gonna want if you can do it. that was our North Star. Our North Star was like, we wanna build something that people want.

Jackson: I remember in the very beginning when we had this list of things we wanted to work on. we were trying out and just interviewing people on the phone and Nate said, oh, we definitely need to do medical documentation. And my initial thought was, there is no way that anyone is gonna let us, like, come in and sit there and watch them interact with the patient and try to like hack together our little toy thing here in order to like make a medical note for them. I thought there was just no chance. I thought it was like dead on arrival and then Nate said, okay, but let’s just get a couple people on the phone and talk to them. this problem was so big. I remember, this is really when the gravity hit me. You know, we were talking to someone that we’d never met before.

We just somehow found them on LinkedIn or something. It was a physician and we were talking about this problem and they said, this is the bane of my existence. If you can come up with software, even if it’s a long shot, that can help me with my documentation, you have full access to everything.

You can come see any patient you want. You can come live in my house. You can come stay in my You know, office as long as you need, like you can come here and just help me, please figure this out. And then I said, okay. Wow. I guess, this is what we’re gonna be working on for a while and we’re still

here.

Bruce Bassi: was, that was the aha moment for you.

Nate: yeah. The first doctor that we worked with, a doctor up in Connecticut and we, had no product. and we were totally honest about this. We’re like, we don’t have a product, we just wanna watch how you do documentation. so we got this like little place next to his office and every day we’d go in the first night we built like the very fir we stay up till like three in the morning, like building like the very first like prototype. It was all running in, I think it was running like in like the code processor, so there was no like UI to it. It was like we would take an audio file, we would copy it into like that black and green terminal that’s in your computer, and then, audio file would get kind of processed in that terminal. Then we would like extract the text, put it in a Google Doc, print out the Google Doc, and then go to the doctor with the out physical copy of the note had no section titles or anything. It was just a, so on the last night we felt like we had something that was kind of good. And so we went over to his house it was late.

His family’s all around, you know, it’s like a chaotic, family home, you know, with kids and dogs and all that. He took out a red pen and he just like started marking it up and he was so excited. You could see it in his eyes, and he was excited about what he was editing and he saw the potential in it. we’ve had, had a hundred moments like that. every single one of those I mean, first of all, we feel super grateful to the providers who have been open to working with us when, they’re not getting like a huge benefit out of it initially, you know, they’re really doing us a favor. but every one of those, like encouraging moments has pushed us forward and kind of caused us to invest more of our selves into this company.

Bruce Bassi: I’m glad I asked that question because I think that’s, a really cool story and, background that you gave about the company and how it got started something I think maybe a lot of new innovators can relate to just being pretty scrappy and, and thankful and grateful for opportunity that you get, you’re first studying out.

Where do you see things going now, I’m sure as innovators you probably have a whole list of, new ways that this can be improved upon. One way I’m thinking is, there’s a lot of, inferences that are made in, for example, our mental status exam in terms of how, the patient is doing to their, tone or, the content, how quickly they’re talking.

since this whole thing seems pretty preposterous that it’s actually exists, is that something that is just a pipe dream that won’t happen for a few years? Or like how close are we to getting to that, sci-fi type of no outcome. Where kind of has read your mind as you mentioned earlier.

Nate: I, I’ll let Jackson answer this, but I will say that tone is something that we need to incorporate into our model. and the reason is because of sarcasm and humor. we’ve gotten this comment on our software a few times now where someone will make like a really sarcastic comment. Our software, doesn’t read tone. That’s just not what it’s built to do yet. sometimes you get some absurd, something like patient is planning to go fly a kite, and it’s like, no, that was an idiom that was used to like, describe them, like going out for the, you know, yeah, tone is definitely something we’re gonna have to incorporate.

dream. For me and for Jackson is to develop five years from now what is a completely keyless EMR, an electronic medical record that, know, has the absolute lowest, kind of profile in a physician’s or nurse practitioner’s life. so everything that we do right now is super focused on creating the best note, but ultimately, like we keep that utopia like at the center of our hearts, like as we continue to develop over the next few years. But Jackson, you know, some of the cooler sci-fi stuff we’re working on.

Jackson: I wanna say to your question, there are a lot of companies out there that are working on, this next level of ai sci-fi stuff that I would say kind of dips into the diagnostic realm. there’s a lot of literature and a lot of exciting stuff going on, especially on just like differential diagnostics and like how AI can help support that. And that’s something that we as a company. At least right now we’re really kind of trying to stay clear-eyed about us not going into that. we see our mission here really as we wanna take this technology and make it so that it can help you as the physician, as the clinician. get all of the stuff done so that you can focus your own mind and your own training and your own background on doing the best, caregiving and diagnosing possible. but we’re not trying to right now get into the space of saying, oh, well, because the patient said this, this, and this, they probably have this, or, there might come a time, the first step that’s going to be made in that direction. is suggested diagnostics that are there just to help. A physician or a clinician,

you can imagine, a note generating, based on everything that was said in the conversation, and then a little side box somewhere saying, Hey, just so you know, based on this conversation, you might want to look into this, you condition that maybe you’re just isn’t top of mind to you. But the data suggests that it could be possible, it’s not saying, oh, this is it, and we’re putting it in the note as like This is what we’re diagnosing the patient with, but some kind of additional support that just sort of. Creatively prompts the provider to think about something that they might not otherwise have.

And this is a conversation that I think is relevant in AI in general. You know, when people talk about using chat GPT for their work, you know, this is something that comes up all the time where people are saying, ah, but it’s not factual. You know? Like it can’t actually give like the real truth. And what people say is, yeah, yeah, but that’s not its greatest use.

You know, the greatest use that a lot of people find Is that it’s a conversation partner to just kind of like help you expand your own thoughts or see things that you didn’t see before, or make connections that were there but you just didn’t quite get to it, surface that for you, but then puts, the decision making power still in your hand. we’re really careful about trying to make sure that we don’t take away any of that power or take away any of that ability for a provider to make those decisions. So over the near term, we’re just going to be laser focused on, nailing the documentation problem. And then from there, maybe even adding some things that just help someone manage their practice, manage the relationships with people, remind you, oh, that patient that you saw a year ago, here’s just like a couple things that you talked about, relevant to, your social interaction with them that can kind of just like help you go back into that encounter. And remember some things. So we have some things like that that are on the near term. But in terms of like the real sci-fi stuff, that’s just not something that we’re like focused on right now.

But I am sure is an exciting field and we just as a society are gonna see, medicine in general change over the next couple years as it gets layered in with more and more of this artificial intelligence.

Bruce Bassi: Yeah, I am really curious as to what the long-term ramifications are to you guys and other companies like yours, saving clinicians, . on a national level, millions and millions of hours. Like, what is going to happen? Are we going to get our lives back? Because I was reading in, Becker’s, said something like 50% of are burnt out and maybe, going to help reduce burnout.

And we’re just gonna say, oh, we got two hours now that we have free for our families, perhaps. Or is it going to be kind of like how . Expanding a highway doesn’t actually improve traffic. More people just end up driving. Maybe the administrators are going to say, oh, you got two hours.

Well, you could do all these other tasks. Now look at that. I mean, , it’s a really interesting concept. And you should have a counter on your website of like how many hours generally, which translates to dollars probably, have you saved across the entire country? It’s pretty cool. I think you are in a position where it’s different than us where we, help individuals on a one-to-one level.

You make a change, I’ve said this in other episodes, changes, improvement is amplified across all of your. Clients and, all of those patients, those clients are seeing. And you can have the opportunity to affect literally hundreds of thousands depending on how large you are at the time.

of thousands of people in the next day, on what an, new feature that you rolled out. Go ahead.

Nate: to speak to the, burnout, issue, the capacity issue and where does the time go? we’ve been shocked at some of the forecasts about physician shortfall. I think that there’s a, I mean, everyone knows about the shortage of and nurse practitioners. but there’s forecast that even in the physician group, we’re gonna be a hundred thousand, 130,000 short in like 10 years. I think a lot of that has to do with, the fact that burnout has kind of this compounding effect, where on one hand it kind of urges people out of the profession early. It also discourages people from getting into the profession. There’s so many exciting things to do, exciting professions to pursue if you’re technically minded, if you are science minded, and in the past, know, a very large portion of that population would go into medicine, and now, that’s less and less the case. one thing that we hope. Is that, we can change the narrative and the tone around how people speak about practicing. again, either as a nurse or as a nurse practitioner or as a, a physician or as a therapist, and, you know, we’d love to see more people actually enter the behavioral health field. I. In general because, I think we all know that like our country and our world needs more, support and needs more research and needs more clinical care, when it relates to mental health. I guess the dream for us is that this is a capacity expanding, tool. That this is a, thing that actually allows our healthcare system to serve more people at a lower cost. and also You know, the mechanism of that is allowing, folks like yourself to spend time with your family on weekends, get home at a reasonable, hour, to not, feel like, pain of burnout.

Bruce Bassi: Beautifully put. Thank you so much for insight. I’ll give you both the last word if you’d like to, Give some closing thoughts.

Jackson: Yeah, we would just like to tell anyone listening, if you’re even remotely curious about our software, go and try it. You can go to our website right now in five minutes from now, have everything set up and be rolling on your one week free trial. there’s no credit card required to try it. If you’re even someone who’s you know, non-medical and you’re just curious about how this works and you wanna see it for yourself, find by us.

Sign up, set up an account and give it a try. You can just go to jot psych.com, Google Jot psych, and you’ll be right there and you can get rolling right away.

Nate: I’ll just add by saying, Bruce, thanks for having us on your show. know, it’s been wonderful to work with you and to kind of hear your feedback on our software, as we’ve been developing it. yeah, just wanna acknowledge that of this would be possible without, providers like you who are willing to generously engage with us and, and teach us how to build something that, is really useful.

So thank you.

Bruce Bassi: Yeah, you’re welcome. This is awesome. A lot of fun. I think what you’re working on is really important for industry and for, physicians and other practitioners on an individual level. And I appreciate you talking about all of these interesting factoids from your company and how the technology was developed to how it works and the impact it’s had on all the clinicians who, use it and their patients because they get to have a more enjoyable visit their clinician.

we will put. the link to jot psych in the show notes and any other, materials and resources that we had mentioned throughout the episode, and we will catch you all in the next one. Thank you so much for being here.

Jackson: Thanks Bruce.