Go back

What Half of Vet Clinics Get Wrong About AI (And How to Fix It), with Sebastian Gabor & Adam Wysocki

from Veterinary Best Practices: A Digitail Podcast

59m 44s

What Half of Vet Clinics Get Wrong About AI (And How to Fix It), with Sebastian Gabor & Adam Wysocki

AI adoption in veterinary practices has grown rapidly, with usage rising from under 40% in 2024 to over 80% by 2026. However, most clinics still struggle with fragmented workflows, manual editing, and poor data integration, limiting real value. While early adopters are customizing AI tools for efficiency, only a small fraction have achieved deep, outcome-driven use. The most valuable applications focus not on time-saving tasks, but on improving patient outcomes, communication, and team workflow. A critical challenge remains data security—over 40% of practitioners fear job loss, and real-world breaches underscore the need for strict policies. Successful implementation requires clear leadership, governance, and a step-by-step approach, such as the six-step AI maturity framework that assesses strategy, governance, and team readiness. Practices are beginning to see transformative benefits, including reduced administrative load and improved client relationships. The future of AI in vet medicine lies in holistic data integration—linking clinical notes, patient history, and communication to deliver personalized care. This shift enables practices to focus on patient care, reduce operational costs, and even support staff raises. Despite this, AI remains a tool requiring human oversight and guidance. Recommendations emphasize starting small by identifying five mundane tasks or broken workflows, then measuring improvements with AI. Ultimately, the key success factor is not just adoption but deliberate, outcome-focused use of AI, driven by empowered teams and clear leadership.

Transcription

10824 Words, 57411 Characters

English
And I would say the big philosophical question is, do we want a faster horse? So notes are done quicker, more rules and the reminders. Or do we want a new paradigm and basically do we want the car where everything gets tailored and the AI does it for you in the background? So if you're spending $99 a month on a product that gives you 30 seconds back, well, that's an expensive 30 seconds, right? If it gives you 30 hours back, that's the best money you've ever spent. We've seen that practices that are relatively perceived as more mature practices get up to six times more value from AI than the practices with low maturity. And basically this bucket, they start now thinking not in features or can I do dictation, they start thinking in outcomes. So stop asking about, can I get another scribe, but more like, can I get home by six o'clock? AI isn't just another thing in that toolbox, AI is like having a contractor standing next the toolbox. It's been a while. Welcome back to another episode of Veterinary Best Practices. I'm your host, Dana Pinduro. After 20 years in clinic, I am on the hunt to bring you the smartest, most talented, best people to help us revolutionize the veterinary industry so we can all elevate our practice. And today I have not one, but two guests. I have Adam Weissaki and Sebastian Gabor. Adam Weissaki has spent more than 30 years building software. He is a founder of Vet Software Hub, an independent vendor, neutral directory, and the creator of PIMS selection navigator. He works with startups, acquires, and established hospitals on PIMS selection, conversion, and workflow mapping with a focus on avoiding expensive mistakes. Gabor is the co-founder and CEO of Digital. He's a serial entrepreneur with over 15 years in technology, Sebastian previously founded IT Gambit, a development studio he scaled to more than 40 engineers across Europe before launching Digital. He is also a co-founder of Code for Romania, which is Romania's largest civic tech nonprofit. So two years ago, fewer than four in 10 veterinarians were using AI, and today that number has more than doubled. The question is no longer will vets use AI. It's what our practice is doing with it, and how can they get better results? In this episode, we're going to walk through the key findings of a 2026 study on AI and veterinary medicine conducted by Digital and the American Animal Hospital Association. Shout out to aha. With Adam and Sebastian here, they're going to help us make sense of these results. We're going to break down how AI use some practices has changed since 2024. Where adoption looks different in the industry, what are some concerns that are still holding professionals back, and what results practitioners are really seeing? We're also going to tackle some topics that people like talking about, how to adopt AI securely, and how to know whether it's actually paying off in your practice. And because, as always, my goal is for this to be usable, we're going to end with some steps that you can take as soon as this podcast ends, so without any further ado, let's jump into my conversation with Adam and Seb. Thank you guys both for joining me and all of our audience here. We're excited to dive into the study that was conducted by Digital Tail. That was just a deep dive in AI use and veterinary spaces. The goal of this podcast is always to give our users something tangible that they can walk away from. So we're going to end with what action items can they take with them today to improve utilization of AI within their practices. Seb, I'm going to start with you. So in 2024, less than 40% of veterinarians were using AI. Now it's over 80%. Did that surprise you at all to find that in the data? I would say first things first, the study that we've done in collaboration with Alhain 2024 was about understanding perception, right? So just like do people are people excited? Do people love AI? Do they want more of it? And the conclusion of that one was that people wanted more, the more they tried, the more they wanted. Great learning at that stage. And now in 2026, the question was like, okay, everyone's excited, everyone's using it. But is it bringing any value? And that was a very, very important question to ask. And I would say the most surprising aspect was that one in three practitioners are building things custom with chat GPT, Claude, Gemini, and all the general AI tools that are out there, which on one side it's absolutely amazing. We're now in an era where people are trying out new technology and playing around with it. And also the fact that another 30% are having the intention at least to build something with those tools. So these two things, very, very surprising. And that they're not, basically, that their neurons are not waiting any longer for vendors to wake up out of their like dream somber and try to do things. So that was also an interesting insight. And I think that we've talked a lot about that we're living in this world where we kind of stopped hoping from our pims or from the current tools that we have to innovate. We're just like, okay, this is the same tool that I've been using for 30 years. I shouldn't expect it to do anything more. So I'd rather spend time to play with Claude or chat GPT. This was the, I would say, a bit of a shocker in the entire conversation. But on the flip side, the problem is that the more you use these different tools, the more the context stays in different tools. And it's not only that the context, the information doesn't come back in one system. But it's also that it doesn't know you. It doesn't know all the records the way you've practiced medicine, what do you care about, and just becomes very, very difficult. I'd say one more thing, surprising, and then, again, happy to answer any other question. But the second thing is that there was one that we've done, basically, a ranking of concerns. What are the concerns people have around the eye? And I would say one that was very high, well, high relatively in 2024, was that 40% were concerned about fear of job loss. That AI will come and replace you. And given that all the innovation, everything we've seen with AI that it does so much more, I would have honestly thought it will be even higher. But surprise, it actually has to almost 20% of the respondents are concerned about job loss, which is a great realization because people are starting to see that AI is not here to replace. It's here to take over all those annoying things that you need to do in a daily basis that you don't want to do. And then you can focus on the things you love. So it's going from will it replace me to how do I run it? How do I use it? How do I leverage it? And also, we've talked about on this podcast, what do you do with the time that's given to you? So are we being intentional about this time that we've gotten back? If we're taking away the administrative burden, are we spending more time with patients? Are we improving education within our teams? Are we taking time to have lunch, go home, see your family? What are we doing with this time that we're freeing up from our unwanted tasks? So given that AI has kind of democratized the ability to innovate in a technology space, how do you help digital, meet the moment where these users need the technology to be and want it to be, but maybe don't know where to find it. OK, so there's so much we're doing on this topic. From AI frameworks on how to adopt it, how to assess your level of maturity, there's really a lot that we're doing, but I would first want to say that there's probably three types of clinics out there at the moment. And there's the ones that are interested in moving away from the chat GPT to tools that are customized and that have already guard rails into them. So what I mean by this is that you don't need to solve the problems of will AI give medical recommendations to the pet parents, well, I'm not there. That's really solved. You don't need to do it again. You don't need to worry about how do I make it sound like me, how do I make it do things the way I want to. All of those are already solved. So people are looking more and more for solutions that are able to do that. The second one is going away from drafting to doing. So what I mean by that is most of the tools will, oh, I need to send a message to a pet parent. It will draft the message for me, so I don't need to think it. But we're going more and more to, if it's already drafting it, and I've seen that it drafted for the last three months in a really well format, can't it just send it to the pet parent? And I just review them afterwards. So this is the second I would say bucket of users that are already quite advanced and they want more from from their tools. And then it goes, the third bucket, I would say, it's also the ones that have seen the most value out of AI. So looking at what's stopping me from getting home at six o'clock and then starting to delegate and more and more to the assistant. Many are happy to talk more about all the frameworks, everything that we've set together, but just to, I would say it's important to make this distinction between the types of practices because at each level, everyone will want something different. - Absolutely. Adam, I'm gonna kick it to you. So in your discussion of the study wrote, use is no longer the story. Management is speaking of job creation. Do we need AI managers in veterinary clinics? If only nearly half of the practices that use AI don't have a plan in place, what are we missing here and how can we make this tool useful? - Yeah, I think that the problem and what I saw in the data here is in almost all the cases, the AI in the practice wasn't intentionally selected. It just started showing up, right? So it was scribes and it was some AI features that started being implemented in the PIMS and it was the practice manager doing using chat to GPT to write social media posts, right? So it just appeared and now it's everywhere and it's being used almost every day, if not every day. So you definitely need some type of policy in a hospital to be intentional about the use and that's kind of, I think throwing your arms around AI and kind of, first of all, understanding everything in the hospital that's using AI and leadership should always know that and have a plan for managing that. And then I like to call it just a one-pager. Like, what can we do with AI and what should we not do with AI? And that might be as simple as like no PII inside of chat GPT or cloud, right? Or, you know, before anything is generated by AI and goes out to clients, have somebody else review it and sign off on it. Every hospital should be able to put that one pager together and make sure the staff is trained on it and understands it. And I think that's the management piece to get started with. - I need to add something on this one. Sorry, sorry then, but like, Adam, you said about the PII. I don't know if everyone's read, but John Hopkins, they had a case, a really huge case now where one of the employees, because no one told them that they can't or cannot do it. They basically use chat GPT or either, I think it was chat GPT in their own personal account. And they passed on 3,000 patients information, like the information of 3,000 patients. And they basically had to now go to everyone and disclose that their data is now somewhere in the cloud with some data, with some AI providers out there. And that was again, very massive and very risky. So it's as simple as sometimes just telling people what we can do and what we can do. That would solve this and prevent all of this. But yeah, without knowing you end up in these situations. - Absolutely. And that's what I think about when I think about AI management in a hospital, and not so much the, you know, the implementation or the technical side of AI, but just how are you gonna use it? And what are those algorithms? - Absolutely. So you, we've talked about this in other scenarios. People pick a software, implementation is an entire horse of a different color. It's not enough to just see the end result of this beautiful demo, but actually figuring out how it fits into your clinic, how it fits into your staff's workflows and making it useful is a separate venture. Does AI add complexity to that or is this the next chapter in a book we've already written? - Yeah, I think it's just the next chapter. It's, you know, the same things apply, right? Number one, you wanna vet the vendors that you're gonna consider their options. I think the one thing that I always like to make sure that we're asking about AI is data use, how their, what LLMs are you using? What are they doing to protect your data once it goes into those LLMs? Very important. Can you get your data out? I think is another one. And you know, this is a, I think a very interesting thing that hospital owners are asking me about more and more often is like, hey, I could probably use Quad to learn a lot from my transcripts of my AI scribe conversations. I had a doctor the other day say like, I wanna find my doctors that are having the most success recommending care plans and find out what they're saying and then teach all the other associate DVM to that script, right? Could I do that with an AI scribe? And I was like, yeah, 100% you could do that with an AI scribe. So, so I think that part is important. And then it goes back to the basics, right? Like what is the problem that you're trying to solve? And pick one outcome that you'd like to see improved. I say gather the metric prior. So if you want, you know, better client, faster response times to client emails and text messages, right? Understand what your response times are now. Implement that AI tool that is promising to do that and then run it for a month and measure it and see if it's improving and see if by how much. And then I think it goes back to what you said, Dana, like with that extra time now, what are you gonna do, right? And also have a plan like that. But I think in the evaluation process, it's important to be intentional. Don't try and solve 100 problems all at one time 'cause you'll fail at 100 of them. Pick one problem, solve that, measure it to make sure and then move on to the next problem. I love that. Okay, wanna take a little bit of a turn. So if this is off limits, just let me know. But I'd love to know both of your thoughts. You mentioned some of your clients add them better, like we want to use AI to analyze our transcripts, figure out how we can do better, figure out how we can prove. That is not unique to them. There are other people who are recognizing the value in veterinary data and we have these AI scribes and AI as just omnipresent in our clinics. Is there something that users should consider or is easily accessible for them to look into when they're bringing that technology into their clinics to protect themselves or just a thought experiment around how that data's being used? - Yeah, I think that is an excellent conversation have with your vendor partner, right? And I call it vendor partner because their job is to bring software in and help the practice be more successful with the tasks that they have to do every day. And once you evaluate the software and understand whether or not it's gonna be a fit for your team and your workflows and it's the right software you wanna move forward, it's a good time to have the conversation about, all right. My data, aside from my physical location is probably my second most valuable asset. And we've thought of software in the past as just an input output and a storage device for my data. But today, my data can inform better operations and I was just at a conference the last couple days and I had a doctor say to me, at the point in time we are in history right now, I can't raise prices. So I have to become more operationally efficient and I'm looking to my data with the help of AI to understand where those places of efficiency can be made. So with your vendor partner, how the conversation is my data secure, who are you sharing with, what kind of systems are you putting it into, and then also, how can you help me access my data with other AI tools so that I can learn and improve and understand what that looks like. And I always say never take it on a promise so if you could get it written into your sales contract, that's even better. And then now you have that new software partner and then you can start working on a plan to leverage your data to improve your workflows and efficiency. - I might add the few things to this one. So transparency as you know is one of our key key values editor and even though sometimes we might upset someone, we have someone like our allegiance is to the doctors. We're independent, we care about the vets and our business model supports that. And I think that's a really, really important question besides the data security, besides everything you own with your spot on Adam, but also with asking the question like, how do these businesses make money? Because if you're suddenly not paying for the product, you are the product. Similar with social media, we're not paying for it. So hence all our data is out there. Meta just did an amazing move. I don't know if you've followed, but every website in the world has pixel installed on the Meta pixel to basically understand how many users go on your website. And if they, if you want to retarget them with ads, you need that pixel. Now Meta just said, all the websites in the world that use pixel, we have the right to take all their data and train our models with it. - Yeah. - We're not paying for it, we become the product. And I think this is very, very important. Every time a vendor comes and says, offer is a huge discount. Anything else, how do they support that discount? How else, what else do they do? Do they sell something else on top of that? Do they actually sell the data somewhere? So I think those are very, very important questions that probably it's not that obvious from the beginning, but if you start asking and go down that, like train of thought, you start getting to those moments. And we see businesses now, even in the vet space that get sold because they no longer have a business model or that was the actual exit strategy to be able to acquire a lot of data and then get bought by someone else. So I'd say that's one important thing. And the second one is, you mentioned it really. The really awesome use case, right, is that you have all the transcriptions, can we now get insights out of that? And I would say that the power of AI is in context and data. If the data doesn't flow well between platforms, if it's not in one place so that you can actually read it, the AI will only give you recommendations based on one single part of that information. In this case, it's amazing to look at all the transcripts, but it wouldn't be even better to understand what was the actual treatment plan if it was actually invoiced and if actually the pet brand came back again, or not, and if the pet is now healthy. The more context you have, the more day I can actually give the proper recommendations and to have a holistic view on top of that. And I think this is one of the big, big questions over the next period is how do we create that holistic view for the AI to be really, really powerful? I bring this up all the time in your episode. I love the context angle because I think one of the great things AI is really good at is sentiment analysis. It's always been very good at sentiment analysis. So when you have an all-in-one platform, when all of your client conversations are captured and scribe AI reception to a texting email, and you have all that context, now we can start identifying client sentiment and we can start pairing clients with team members that are better at handling the fractious clients or the clients that get overly emotional to not only have a better client experience in the hospital, but also to take a lot of the stretch of pressures off your team by just having personality mismatches in the exam room. So that's a great context of super important. Yeah, we actually, I've worked with a group here at Digitale that's doing just that in such a beautiful way. So they've trained part of the AI to grab sentiment, but also grab communication style from their customers. It gets tagged in the pins and they actually have different discharge notes and an entire communication cadence set up with them based on the studies that they've done and they have kind of sectioned out so they have people that would like a lot of data, people who are more emotionally connected, but every single employee has this information, has access to this information and then has a basic guideline of how to connect with that person so that compliance increases, they feel heard they're getting what they need from that partnership with that vet. No one had the time to do that before you would have had to pay someone to stay in the room and, you know, like, scribe and put sticky notes everywhere. This is something that wasn't possible even five years ago and there it's changing the way that they're connecting with pet parents. Absolutely. It's super excited about how we're going to see more issues like that evolve over the years ahead. Oh, yeah. It's fun to be on the cutting edge of that. So kind of going into, yeah, go ahead. No, I would add one more thing because you just said this and there was one of the, you know, one interesting insight from the, from the studies that 61% of the practices still send rule-based reminders and only 11% actually tailor all those reminders and all that communication for the patient. So we're just scratching the surface in terms of what this, what you can do with this and how many people are adopting it, but it's very, very powerful. So notes don't quicker, more rules and the reminders. So basically that's, I think, the paradigm shift that we're solely getting into at the moment. And that's, I think the economy is kind of forcing our hand on that. If you looked at during COVID, we had more business than most people knew how to handle and wanting to take care of as many patients as possible, just so everyone was seen. And a lot of the push around the technology at that time was how to improve operations from a speed standpoint. Things are slowing down for many clinics and client retention and connection compliance become more prevalent. So one of the things I think about as someone who cares for pets is, how do I educate my clients enough to know when to come in? I could have done, you know, antibiotics and a little bit of pain medication three days ago, but now it's a foldable and abscess repair. So when they're venture to save themselves a little bit of money, they, they've missed a risk. I don't have time to call these people. So having that relationship building, having that connection, being able to get information that's meaningful in front of those clients is what is necessary to navigate tighter budgets, a little bit of hesitancy to come into the, to the vet. So I'm happy to see these AI tools kind of pivoting towards that or people's use case around that pivoting towards that because I think that's a very meaningful improvement that we can make for the quality of our patients lives. I'm always the optimist, but I will have to preface like something on, on this one is that even though adoption has increased tremendously. So adoption, yes, it went from 410 clinics using AI to now almost 810 using AI on a daily or weekly basis. However, if you look at the depth of their adoption, we're, we're still at 16% that have currently built AI native flows built with ender pins and then you expand it a bit more with, with other tools, but you still need to copy paste, you still need to edit at the end because it's not how you wanted it to or it still doesn't have the right context. So we have so many great challenges ahead of us to help more clinics with and but still we're just scratching the surface for, for a lot of what this can do. Yeah, well, thank you for team this up to me because yeah, I wanted to dive into the numbers. So we have a lot of adoption, but you mentioned earlier on the clinics that are using it well are seeing six times more output, but those are not the majority of clinics and as a matter of fact, 56% of these users are spending extra time editing the AI output. I think 45 of them are moving data between different tools and copying and pasting data by hand. So how do we improve AI maturity within a clinic? So you talked about those three levels of users. How can we help everyone level up to a place where AI is working for them rather than creating extra work? So two answers for this one. One is I'll describe in a second as very, very simple six steps framework that any clinic can do, but first I want to give a short story and they now don't kill me if I give too many details from from digital and how we how we do things. But at the beginning of this year, so we were part of a startup community and basically we were looking at everyone like what's everyone doing with AI? How can we be better? So this is more like for us as digital as a company and a team, how do we become better? And the most conversation was that hey, everyone's using it, but it's very chaotic. And I was feeling this also within digital. We have amazing people. Everyone started using everyone was trying their own tools, different frameworks. And the big question was how do we as a team learn from all of this? How do we do things cohesively? Yes, maybe sometimes you won't pick the best tool, but you'll pick the tool that everyone can use and learn from it and share ideas, exchange and become better as a whole. And that was the moment where with a bit of help, like we've also reviewed all the frameworks out there, how to best implement like change management at scale for AI. And we've done a very simple process for us internally to use. And I'll share if you want more towards at the end, like what are the concrete steps. But for medical practices, we've done a six step, very simple process that any vet clinic out there can adopt. Again, we'll have the link probably in the end of the description. You can go online. It's a very quick assessment that will help you understand where you are and what are the next quick steps for you to do. But what's important is that they're six, right? And the first three are around leadership. And the first one is basically strategy and intent. Super simple words is, is there a destination on your GPS or are you just driving around and seeing what's happening? So for that one is you need to have documented path of where do you want to go, what do you want to do and exactly what Adam was saying is like, what's the outcome? What is the outcome that we want to achieve? The second one also very, very easy to get rolling is governance and oversight. Basically write the rules on paper and name someone as the adult when it comes to AI, who makes decision when it comes to AI in our practice. And the third one is, and we have the master here, is vendor evaluation and selection. How do we decide this? Do we just pick the first car in the lot when we go by a car or do we actually have a checklist, compare multiple options and do it in a structured way? So all those three are with leadership. There's so many steps under, under them that can be done, but very, very simple steps. And the third piece is more enabling. How do we enable the team to do more? And the first category is workflow and technology integration. So is it within our tool stack, is it within our workflows, is it connected? Do we stick with the same workflows or do we adapt our workflows for what technology can do? Because there's a saying that at least stuck with me is that AI just accelerates processes. So if you have a broken process, it will just accelerate the results of that. The second one is adoption depth, and this is where, is it basically, do we have just one doctor using it, or is it the entire team using it? Are they using it at the same level? So basically it's more AI than norm in our practice, or is it an exception that we try sometimes to do it. And there's again, very easy. things that can be done on this one, but you'll get a lot of impact by creating a leveling system for the entire team. And the last one, I would say capability and readiness. So is the practice equipped for this? Do we have training? Do we have some budget aside to invest in this? Do we have, like, we're dreaming about AI and what AI can do, but most of the practices still don't have a microphone on computers. So there are still some steps there to make sure that you can actually leverage this technology. And we can talk about all the cool cases that we've seen where they have microphones from the ceiling or just use your phone or Apple Watch. There's so much you can do today. But still, you need to make sure that you've done that. So again, six steps. Very simple. It takes 10 minutes to at least get a sense of where you are and what are the next steps and any practice can start leveraging this. Yeah. Excellent. I mean, I think from my standpoint, when I talk to hospitals, you know, I talk to them about, and we use the word tech stack a lot, I always say, like, tech stack really is like means toolbox. In your toolbox, you have a hammer and a screwdriver and a tape measure and, you know, and AI isn't just another thing in that toolbox. AI is like having a contractor standing next to the toolbox. And do you want that contractor? You want to pay that contractor to just hand you the hammer when you need it or hand you the tape measure when you need it, or do you want them to build something or fix something? And I think that's the basis that I like to see hospitals think about when they think about AI. So not just the mundane tasks or the time-saving tasks that they're going to leverage AI to help with, but what are the broken things in the practice? Where are the communication drop offs? Where are the things that staff is forgetting to do? Where are the things that don't get done that would, but you don't care that much because they're nice to have? Now they can, now they must have, right? Because it's so easy to do them. Start by documenting that list and then that becomes the basis for your plan and you can work with your software partners, your pens partners like the team at Digital Tail or whoever you're using and your other AI scribe or AI tool partners and say like, these are the things that we'd like to fix. Do you have best practices with your tools for this? And then if not, what are the other resources you can use? Well, you can go into a cloud or a chat TPPT with privacy settings on, hopefully using a business account please, and start having a conversation with it. Give it the list. Give it the set of tools you have and then say like, what are some strategies I could use to make sure, you know, recheck recommendations are getting communicated through everybody in the practice properly so we can get those patients back in. Or we, AR, we have a script for when we have to call people for AR and this is a script we're going to follow and these are the options we're going to offer them. You can have AI help you craft these workflows that everybody kind of, they're after thoughts sometimes or pushed off for other things but tools can do that and then eventually a lot of those will be even like client outreach will just be automated. I think that, you know, AI voice automation is still a little early, but it is only going to get better. Absolutely. So kind of take you, oh, yeah, you mentioned one thing, I don't know, I think it's worth to click double, double clicking on this one as you mentioned that again, whatever tool you use go to the privacy settings. I have this conversation so many times so just to clear out for everyone is like when you use chat GPT right because normally the first step is just use chat GPT use something that's easy to try out. When you go there in the settings, there's a setting called do you want to improve the model by sharing the data and that comes by defaults active. That's the one you need to uncheck to make sure that the data doesn't stay with the model. So very important piece for everyone if you're about to try and embark on this journey, that's the toggle you need to check out so that you make sure that you don't give data to the models. And don't feel bad about turning that off. Those models have a lot of money to improve themselves, they don't need you. And by a lot of money, I could add maybe a bit of a very interesting news that just came out. So when we launched the setting, this was like what a couple of months ago, the all the industry insights were that the AI in this room will grow to 1.3 billion in the next five to 10 years. So surprise for everyone on Tropic, which is the second basically the top two together with OpenAI and Tropic just announced that they're going to file for IPO at the two trillion valuation. That's a number that has a lot of zeros that I still need to put them on paper to understand how many they are, but they're a lot. So the AI bubble has not burst yet, that's what we're hearing. 100% no, just growing stronger. Once the whole period on the anthropic shares go off, it will burst. So let's put this in numbers that might be more applicable to our audience. The average veterinary clinic is spending $417, just $417 a month on AI. We've talked about how they can improve it, the ways that we can use it. How do they know they're getting their money's worth? Well, measurement, right? That's the most important thing. And I think in terms of software management and smaller independent practices don't have the benefit of IT teams, a lot of the larger corporate groups have, or the advanced ER specialty. So these IT teams measure, right? What are our processes today? What are the problems we're trying to solve? They understand how many timings and sometimes even down to a number of clicks. So I think when you're implementing new tools, or when you're bringing that contractor slash AI into your hospital, it's important to have a baseline first. And that document of like, what are the frustrating things? What are the things getting lost? How long is the average callback taking, you know, build a plan and pick one to improve? And then measure it and use the new tool and actually see if there's an improvement there. And by what percentage, right? So it's important to have a baseline and then to measure and then really understand the ROI at that tool. So have anything to add? I just 100% agree with what I would probably expand it. I would probably expand it a bit more to, there's also quality of life. So basically, are you enjoying more your day? Are you feeling that you could delegate the things that you didn't want? Maybe it didn't give you a lot of time back, but at least you now enjoy the things that you want to do and you don't need to do the things you didn't want to. That's probably one I mentioned. I can tell you from career experience since I started the business is if a software tool comes in and it reduces quality of life in a veterinary practice, it does not stay long. It does not stay long. That is true. The reception team alone, it doesn't pass much. The subscription doesn't get canceled, but they just stop using it. That's, they just pay for it to just kind of be inside. Which is another great point around these AI tools is like, since a lot of them are really easy to download and get started with and kind of the subscription kicks in, you might not even know because you put a credit card in for a trial and then forgot about it. Every 90 days, audit those tools against what the practice is paying for and make sure that everybody on the team is using them or somebody on the team is using them and getting that benefit out of it. Otherwise you're going to be paying probably a long time before your bookkeeper says like, "Hey, what's this? Do we still need this if they even ever do that?" I feel like that's a mental shift for me with adoption with AI was I kind of just thought, "Oh, I'd create something to be one and done." And that maintenance, so every 90 days, are we using these tools, are we using these tools the way that we want to be using? How should it shift? However, a gold shift can't meet us there, it's a Durham case, guys, it's ongoing and the best Durham case. I mean, but all tools are like that, right? Like at least once a week, I talked to a practice that we're talking auditing tech stack and they've been paying for Canva for three years and they used it one time, right? For a social media post, two and a half years ago. And it was really beautiful. It was really easy. For a conference and they're like, "I'm going to have the best social media in the world as soon as I change my entire business." Right. And maybe we can cancel one subscription. Not looking at anyone in this room who's got that. So take us in, it's a typical Tuesday. We have two clinics. One has a high level of AI maturity. How are they starting their day that first hour and how's that differ from a clinic that maybe isn't utilizing AI very well? Yeah, I think from what I've seen is it's the overnight summary, right? That's my favorite. What emails have come in? What phone calls have come in? What has to be triaged now versus I can get to it later? The men and women that sit at those reception desks and have to go through the tons of messages that come into the practice between, it's crazy. The number that come in between when the practice closes and opens the next morning. An AI that can summarize and categorize those things and even better that can, you know, agentically and autonomously take action on those things and route them to the right people to be resolved versus like, you know, you're listening on the phone and transcribing notes on a notepad and you're manually responding to every single email and thinking, wow, what should I say to this client? The time difference between that A and B practice is incredible. - And the mental load. - If only there would be a solution for that. - Well, I'm vendor neutral, so I can't name the solution, but said, you know. - There might be one. - No, there might be one or two. - There might be more, but yes, at least one that we know of. But I would say for this one for the ordinary Tuesday I love the question. I would say there's two also two things. One is, is the team learning from one another? Are they sharing, oh, I've tried this, this works, this doesn't work. That sharing culture, it's very important when you see it in more mature practices that there's a lot more communication around this. And the second one is, you've very touched a bit on this is that do they have that milestone of reviewing workflows? Are they changing, are they auditing the workflows? Are they checking that they're getting the value out of it? Because the reality is that AI is exactly as your new vet graduate that is at the beginning and maybe needs a bit of support, maybe needs a bit of guidance. You can't just let them do everything, maybe they need to learn a bit from you how you do things, how you want to do things. So it's exactly that with AI and I'm seeing a lot of practices that try it once. It's not the result they wanted and they stop. Instead of tell it to these ones, how you want it to do it and then you no longer need to do it again. That's probably one of the magic with AI is that if with people you constantly need to coach to get the behavior, at least with AI, you coach once and you get the behavior that you want for and it stays there. So, but there are these differences and I would say more mature practices do this while the ones that are the beginning they hope things happen. - Yeah, and I would just follow up. Like the vendors that are supplying these AI tools, this is your partner in solving these things. So if you try it once and you don't get the outcome that you expected, that's an opportunity to reach out to your partner and learn and connect instead of just giving up. So, it's really important when you're evaluating to dig into that and how much will you be there for me? It's okay to say these tools are all new to me, but I want to leverage the capabilities that they're bringing to make things better for clients, my patients and everybody on my staff. So, make sure you're working with somebody that is going to be there for you and kind of help provide some guidance through break podcasts and studies and blog posts. - These resources, yeah, I would say my, I think the worst case scenario is not the person who tries once and it doesn't work in the quit. What hurts my heart is the people who are using it, but they're making manual corrections. It's like, I always want it to record a BCS, score a 509 unless they say, oh man, think of like doing that every single day when you have a system that can do that for you, but you have to have that like that tenacity in that sharing culture, who's doing it really well in your practice? How can you elevate amongst that? And that might not be your traditional leader. So, that's something that I found really fun about implementation with AI and clinics is it's like, oh, they're not the team lead. No, no, no, no, no, but they're an innovative person and an excellent communicator. Or maybe that person who's like a little bit rough and short, that's a good AI prompt 'cause they're gonna tell you exactly what they want. And so, it's not always like a top down thing you were talking about, we use these tools for everyone to be able to create what they want to they're not waiting for our technology leaders to create it. But even within the clinic too, it democratizes that ownership of these outputs and you get just a wealth of knowledge from different people and different expertise. So, it's like, make it, don't delegate practice managers. I'm gonna say this all the time. Please delegate, please include your team. - Yeah, and Dana, I think that's another important point that I always like to bring up with hospitals is make sure you're making space for the people who are getting frustrated or have questions so that they feel comfortable being able to ask their peers or ask leadership or help and assistance on learning the tool. And then those people who are super helpful and giving the tips and sharing with their peers ways to accelerate, like recognize them. And maybe as you're implementing something new, especially if it's a complex system, over the first few weeks, do some gift cards. Hey, number one helper, you know, number one tip of the week and make people feel seen and it will pay back itself tenfold easily. - Absolutely. Okay. - I was a few weeks ago at one of our practices that's using digital and they're five doctor practice and we were doing a launch learn on like use of AI 'cause they were at the beginning of the journey. And one of the doctors who hadn't tried it before was like, okay, but we'll be able to do, to basically pick up on all my short snippets that I always like write down. Will it understand that I wanted to write this? And even before I started saying anything, one of the technicians jumped and answered, oh yeah, I do that all the time. - I love that. - Yeah. So the knowledge is there within the team, just empowering the right people and delegating to them. - Well, because it's in them, like, you know, we're providers around the outside looking in, but their ideal flow, their ideal connection, their ideal output comes from them. So without that investment and participation, you're just regurgitating what someone else thought you might like. And the AI allows you to customize it. So we don't have to have these prefab reminders that just sound like, you know, thank you for calling. You can create something that sounds like your team, that sounds like your character, that is a reflection of the workflows, the efficiency, the inefficiency, whatever it is, that's your special sauce. That's what it allows us to do. But that comes from you, that can't come from anywhere else. That's not a little like care bear, but. - I think it's great. - Yeah. Okay, so we do the study again in two more years. What are you guys hoping to see? - I'm hoping to see that the management side of it grow. Like, practices really taking control of these AI tools in a way that steers them to help make improvements. And, you know, I think that one of the things a lot of hospitals struggle with is improving workflows of a time and getting a staff of 20 or 30 or 40 to all kind of move in cadence into a new workflow. And I think with the AI tools that are out now and that are coming and will be improving in two years, with some smart management, some plans for how to implement and how to, you know, run AI in your practice, we can get workflows really, really dialed in and specific to the practice, to the teams within the practice, to, you know, lower the operational cost. And hopefully, you know, get our arms around some of the pricing pressures that we're getting in hospitals so that instead of having to raise prices, we can keep prices where they are and then pay staff more because we're gaining so much operationally. - I love that. Raises, that's what we want to see in two years. - Raises, that's right. (laughing) - So what are you hoping to see in two years? - Two years, that's a long, long time to predict. But if there's one big thing I would love to see more and it's related to depth of adoption, but I would love to see an instance of what Adam was saying. I would love to see more clinics adapting their workflows to today's technology and what they can do with it. And one simple example is that we've built and vet care more around these snapshot moments where you come at the practice every once in a while and that's the moment when you learn what's new, see what's happening and get advice. And the reason why you do that is that there's no vet or there's no staff that could afford to call every single pet parent every single week to check in on them, give them the advice for this age and what's the risk for this breed, the breed, the sage, this given this history. But the thing is that AI can. And the cost of doing that with AI is so small that you can actually afford to give it, to give this type of experience to all your users, all the clients you have. And even more is that you would then maybe ask, yes, but what if I suddenly come in the morning and I have a thousand messages, I won't be able to read them? The good news is that AI can also triage those for you on that. And this is one of the examples that we've already built and we're seeing more and more clinics taking advantage of this, but it does require a bit of courage to think, how can I provide vet care today to all my patients in this new world and this new context. So the more examples like this, I would say the better and that would be the hopes for it to end two years from now. - I like that. And for, I know you said two years is a long time, but for any AI skeptics out there, how should we best prepare for AI overlords to take over the world? - Yeah, be nice to your phone. That's, be super nice to your phone. - Yeah, I think we're just at scratching the surface on, I shift towards autonomy, right? And these agents kind of, I just traveled to a conference I used, I was playing around with news, found from Meta, it booked my whole trip for me, right? It booked the flight, it booked the hotel, it scheduled my ubers, like I had a real personal assistant, gave me three options, I picked, paid for everything. So. This is the future that's coming and I think that we're going to we're going to start seeing and that we're actually already seeing is agents calling hospitals asking for appointments and availability and pricing information And I just picture that times a thousand over the next two years You know my TV is going to be talking to my washing machine. Please pause right now. Adam's watching a Marvel superhero movie That's the dream. Yes, yes Less time doing laundry and more time spent with Marvel 100 percent. That's the AI dream. So I do That's why we did it So do you have any any tips for surviving the AI takeover? Yeah, this this reminds me of one of the conversation I had with a vet friend a couple of weeks ago and he was saying that look I have patients now coming in with plans made But with chat GPT and honestly what I just want to do is give it to my AI really I just review it and then send it back So as you said like AI talking with AI sorting things Amongst them, but so I need to flag this I am a techno optimist in general So I do believe more technology will help and empower more people to do the things they love And we're seeing this already I my biggest prediction for this is that every team We'll have their own assistant to help them and take over workload from each team So and we're seeing this already this direction with how we're also building Tails, but you now have a tails concierge that's helping the the front staff team. You have a tails medical helping the the medical team And tails practice major helping the the practice management team But when it comes to robot takeover, I would say first we're very very far away from that one because a robot can't hold a scared cat So until we get optimist and everything else from Tesla getting to that level There is very very far in the future, so I would not worry about that But I would say that The biggest outcome that we've seen also with the study is that There's more time for patient care and I would more focus on the fact that this is what AI can give us today Not in the future. It can really do that today and we're already seeing it across the industry And the rest I honestly we're building with AI and we know how difficult it is And it needs so much coaching that I would love for it to suddenly become 10x smarter and to do things on its own But it's not there. It still needs a lot of guardrails. It still needs a lot of like human intervention And that's also why we've broken down Level of usage of AI into like three big buckets, but anyway, we can talk about that and other things I love that you're metric for AI takeovers if they can handle a fractious cat Like humanity is safe Listen Grappot is never going to get a blood draw out of my 18-year-old kitty. She just But my vet will here's the here's the thing to think about and I've been in the veterinary industry 10 years And I think about this all the time for 10 years I've been hearing how you know veterinarians are great with patients and clients not so good at client communication and running businesses and all the That stuff. Well, AI is going to do all that for you, right? And so like think of all that that frees up for patient care and access to care more doctors are able to open up hospitals because they can have an AI Agent autonomously run the nerd stupid stuff about the business and they can just focus on Taking care of animals, which they've been dreaming about since they were eight years old, right? So I think that part is is really really cool and nobody should be afraid of that No, that's great. I always say like the biggest sad day in a vets career is when they realize there's an owner on the other side of the leash and they're like Yeah, I've done We're like like who in the building wants to do inventory nobody, right? Especially like when the kitten know when that when that when the when the litter of kittens comes into the exam room and like you're gonna count pills No, I don't think so. You the kitten goes on your shoulder while you count pills That's the dream. Can we spread sheets? They can let we'll let uh Elon's robots count pills and take care of the case Okay, uh final question After whoever's listening and furiously taking notes is inspired turns off this podcast. What do you want them to do first to start implementing change? Oh, I want the lens of the AI conversation. Yeah, it's really not changed too broad. Don't take it too big Adam will start with you right right down, you know five five mundane things in the practice that could be replaced with AI you know Hopefully digital things not like mopping the floor although that's coming um and and then you know five Five workflows that break frequently in the hospital, right? And then Set out a plan to see if you can address those with AI through your software partners or through doing research on You know great websites like that software. How calm? I don't know if it's a little self-promotion there Because I think you would find a lot of that is fixable and will free up time and resulting better patient care Through because you adopt AI Yeah, I love that So what's the first thing they're going to do when we're done with this podcast? The simplest version is honestly go to our AI maturity framework and go through it It's going to give you all the steps and next up depending on where you are today Because if you're at the beginning of trying to use anything with AI It's a complete different next steps that you need to do rather than if you're someone that's already using it And you have a team that's everyone's using different things So I would say that's the most practical thing you can do you click on a website spend five minutes go through the assessment And you'll get the exact next steps But if you want to concrete like now two things is one Name an owner so that's part of the governance name and AI leads in your team And like Dana you were saying it doesn't have to be your practice manager It doesn't have to be any team lead it just whoever is the most passionate and curious About this they're the person that just give them the empowerment to go and find out what's new And let them come back with ideas and always try to and then follow Adam's advice like find the outcome work back from that But name that owner because that's the biggest blocker and the biggest difference between practices that succeed and practices that don't um Yeah, there's so much more, but I would say that's the that's the first step that's the first step Yeah, okay, wait no, okay, no Um, I would say there's four four levels of company um There's a AI adoption and oh lights went out. Okay, uh, so one is the one way. I'll ask your music. It's playing Um, so four levels of company and regarding the AI maturity and the first one L1 is where most of the people don't use AI So that's L1 if you if you think that's where you are today. Well, that's fine But it's it's an easy. It's the first step L2 is most people use it, but the use is sporadically Different tools different ways not automation. It's more like think of Instead of going to Google you now use Chagypity to search think that's basically L2 type of approach L3 is where AI takes over recurrent tasks So if you already see that there's some recurrent tasks that AI is doing Think of AI receptionist picking up phone calls think of AI refills all the process behind the scene. It's done behind you just come and approve it or or not So if you have those type of examples, it means you're now coming closer to L3 as a as a company And L4 is probably add on favorites is when AI owns outcomes In the practice And that's when you now have a full AI strategy you treat them as a new team member That's helping all your teams and taking workload from you But I would say these are the four levels and it hopefully it's it's a bit of guidance to understand how much more you can grow on the On the scale and then again for next steps definitely 100% I recommend the Framework Figure out where you are now then take that first step forward Okay, I want to thank you guys both for joining me and for this discussion And in two years, I'll have my AI agent book something with your AI agent so we can review the next study And all of our agents will be on a podcast together You know Yeah, but they won't be able to wrangle the fractious cats and that's the important part What I loved most about this conversation was how practical their advice was if you're wondering what next steps You can take towards great AI adoption in your clinic here are three things to start with First name an AI owner Pick the person on your team who's the most curious about it and give them the room to lead that transformation Second, you're going to write a one-page policy on what AI can and cannot be used for Starting with no patient or client information and the tools that aren't approved Third, you're going to pick one problem you want to solve with AI Measure it before you start and measure it again after a month so you know if this tool is earning its place If you're not sure where to start or how your clinic is doing with AI check the AI maturity framework on digitale's website Takes about 10 minutes and it's going to give you practical next steps for your setting the link is in the description Thank you guys so much for watching I'm excited to hear what you have to say in the comments and I'll see you next time on veterinary best practices You

Podcast Summary

Key Points:

  1. AI adoption in veterinary practices has surged from less than 40% in 2024 to over 80% in 2026, shifting from curiosity to active use.
  2. A significant portion of practices are now custom-building AI workflows with tools like ChatGPT, indicating a move toward hands-on innovation.
  3. Despite high adoption, only 16% of clinics have built native AI flows, and many still spend time editing or copying content between tools.
  4. Practices with higher AI maturity report up to six times more value, primarily by focusing on outcomes rather than just features.
  5. AI is transforming workflows by reducing administrative burden, enabling better patient communication, and freeing time for client care and team development.
  6. A key concern remains data privacy and security, with real cases of data breaches highlighting the need for clear policies and vendor transparency.
  7. Successful AI integration requires leadership oversight, clear governance, and a structured framework to evaluate and measure ROI.
  8. The future lies in holistic data use—linking patient histories, notes, and interactions to provide personalized, context-aware care decisions.

Summary:

AI adoption in veterinary practices has grown rapidly, with usage rising from under 40% in 2024 to over 80% by 2026. However, most clinics still struggle with fragmented workflows, manual editing, and poor data integration, limiting real value. While early adopters are customizing AI tools for efficiency, only a small fraction have achieved deep, outcome-driven use.

The most valuable applications focus not on time-saving tasks, but on improving patient outcomes, communication, and team workflow. A critical challenge remains data security—over 40% of practitioners fear job loss, and real-world breaches underscore the need for strict policies. Successful implementation requires clear leadership, governance, and a step-by-step approach, such as the six-step AI maturity framework that assesses strategy, governance, and team readiness.

Practices are beginning to see transformative benefits, including reduced administrative load and improved client relationships. The future of AI in vet medicine lies in holistic data integration—linking clinical notes, patient history, and communication to deliver personalized care. This shift enables practices to focus on patient care, reduce operational costs, and even support staff raises.

Despite this, AI remains a tool requiring human oversight and guidance. Recommendations emphasize starting small by identifying five mundane tasks or broken workflows, then measuring improvements with AI. Ultimately, the key success factor is not just adoption but deliberate, outcome-focused use of AI, driven by empowered teams and clear leadership.

FAQs

While less than 40% of veterinarians used AI in 2024, over 80% now do. The shift is from basic usage to deeper integration focused on outcomes, not just features, with practices increasingly using AI to improve patient care, reduce workload, and enhance operational efficiency.

Practices should establish a baseline of current performance (e.g., response times, callback times) before implementing AI. After deployment, measure improvements in key metrics to assess ROI and ensure time saved is being used for meaningful activities like patient care or team development.

Level 1: minimal or no AI use. Level 2: sporadic, ad-hoc use across different tools. Level 3: AI automates recurring tasks like reminders or note drafting. Level 4: AI drives outcomes, with full integration and strategic use to improve workflows and patient experiences.

Clear governance ensures data security, prevents misuse of patient information, defines roles (like an AI lead), and establishes policies—such as requiring human review before sending AI-generated client communications—to maintain control and compliance.

AI can analyze client interactions to identify sentiment and communication style, enabling personalized outreach. This helps match clients with appropriate team members, improves client experience, and reduces staff stress from handling difficult conversations.

Patient data can be exposed or shared without consent—such as in the John Hopkins case where employee misuse led to 3,000 patient records being leaked—so practices must disable data-sharing settings and enforce strict data privacy policies.

Chat with AI

Loading...

Pro features

Go deeper with this episode

Unlock creator-grade tools that turn any transcript into show notes and subtitle files.