AI Shay: A Beginner’s Guide to AI in Facility Operations

Artificial intelligence can feel overwhelming, especially for facilities professionals who may think they don’t have the time, technical knowledge, or need for it in their daily work. In this episode of Healthcare Facilities Network, Shay Rankhorn breaks down why getting started with AI may be easier — and more practical — than many think.

Designed for the facilities leader who may be skeptical or unsure where to begin, this conversation focuses on simply starting. Drawing from his own experience in the same seat many listeners sit in today, Shay shares approachable ways to begin using AI to support everyday tasks, improve efficiency, and make work easier — without overcomplicating the process.

For listeners already experimenting with AI, Shay also offers practical tips and tricks to augment your current approach and get even more value from the tools you may already be using.

Whether you’re AI-curious, AI-cautious, or already testing its potential, this episode offers practical guidance to help facilities professionals take the next step.

🚨 Subscribe to the Healthcare Facilities Network Podcast to gain awareness about the rewarding career of healthcare facilities management: / @healthcarefacilitiesnetwork

👥 Connect with Peter: https://www.linkedin.com/in/peter-martin-6284363b/

Today's Guests:
👥 Connect with Shay Rankhorn: https://www.linkedin.com/in/shayrankhorn/

✅ Important Links To Follow:
👉 Cref: https://cref.com
👉 HFN Website: https://healthcarefacilitiesnetwork.com/
👉Watch on YouTube: https://healthcarefacilitiesnetwork.com/
👉Listen on Apple Music: https://podcasts.apple.com/us/podcast...
👉Listen on Spotify: https://open.spotify.com/show/3vWorrUkfsBprrg5S2TUjZ

✅ Stay Connected With Us!
👉 LinkedIn: https://www.linkedin.com/company/healthcare-facilities-network/
👉 Instagram: https://www.instagram.com/healthcarefacilitiesnetwork/
👉 Facebook: https://www.facebook.com/profile.php?id=61572807523974
👉X: https://x.com/HFNPodcast

📬 For Business Inquiries: pmartin@cref.com
=================================
Disclaimer: We do not accept any liability for any loss or damage which is incurred by you acting or not acting as a result of listening to any of our publications. For all videos on my channel: This information is for general & educational purposes only. Always consult with an attorney, CPA, or financial professional for advice based on your specific situation. Copyright Disclaimer: Under Section 107 of the Copyright Act 1976, allowance is made for "fair use" for purposes such as criticism, comment, news reporting, teaching, scholarship, and research. Fair use is a use permitted by copyright statute that might otherwise be infringing. Non-profit, educational, or personal use tips the balance in favor of fair use.

© Healthcare Facilities Network.

Transcript

The PBJ Sandwich Prompt Lesson

SPEAKER_01 0:00

The analogy that I use today that I use all the time is think of asking your four-year-old to build a peanut butter jelly sandwich. You don't just say to a four-year-old who's never built peanut butter jelly sandwich, hey, go make me a peanut butter jelly sandwich, because what you get back may not be anything resembling that. You have to tell them what is jelly, what is peanut butter, where is it located, how do you get the lid off, where's the bread, what type of bread, how do you get the wrapper off the bread, you know, do you use the heel or not? Uh, and then where's the knife? How you use the knife? Do you put peanut butter on both slices of bread or just one? You know, all those different things that you would teach a four-year-old, if you go through that same mindset as you're teaching your AI just to talk through, uh, you can give engineering prompts that are very crystal clear as to what your outcome is. That's one way to do it.

Shay’s Shift From Hospital To Consulting

SPEAKER_00 0:53

There's a major crisis facing healthcare facilities management. We have aging employees, aging buildings, and aging infrastructure. We've created the healthcare facilities network, a content network designed specifically to help solve for these three pressing issues in healthcare facilities management. We bring on thought leaders and experts from across healthcare facilities management. All the way from the C-suite to the technician level, because at the end of the day, we're all invested in solving the aging issue. Thanks for tuning in. Look at our videos, you will find that is a theme across our content. This is the Healthcare Facilities Network. I'm your host, Peter Martin. Day one, Spring Conference, Allen, Texas. I'm happy and pleased to be joined by Shay Rankhorn.

SPEAKER_01 2:01

Hello everyone, I'm Shay Rankhorn. I'm a partner at Facility Diagnostics for a commissioning firm that works all over the U.S. providing high quality, quality control, and uh energy reduction projects for clients uh through multiple modalities or service lines, whether it's healthcare, hospitality, government, higher ed, and so forth. Prior to that, I had a 27-year career in healthcare and I was the 2022 president of Ashie. And I'm glad to be here today with my grandvater and uh have a good conversation.

SPEAKER_00 2:33

Shay presented the um afternoon keynote talking about AI, artificial intelligence, but I thought he really broke it down into digestible, but more importantly, practical, practical ways people can use it. So that's why I'm really happy that Shay is joining me. But I wanted to ask you, Shay, you've gone from vice president of hospital facility, a system, all the challenges of that, now Parker facility diagnostics, you've been here for three years, so you've jumped over the consulting side. How's that been for you? How's that transition out of the daily war into another daily war gone?

SPEAKER_01 3:10

Well, that's an accurate way to describe it, going from one war to another, because when you're working for a company, um you feel whether rightly or not, you feel somewhat assured that that paycheck is coming in every week. It's more steady and stable. Whereas when you own a business, you realize that you're out there having to constantly go source work in order to feed the 11 families that depend on FD. So it's a different level of stress. But the beauty of it is now is I get to guide and direct the ship, and I get to ensure uh that my employees have the last job they'll ever work because they love working for us so much, and I don't make that up. My employees tell that to our interviewees because that's what I tell my employees all the time. Tell me how I can make this the best job you've ever had. What can we do to ensure that you own the process, that you have a say in the outcomes, that you feel fairly compensated, and that we adjust for family and other issues that come up? Because if our staff are not taken care of well at home, then they can't function well at work.

SPEAKER_00 4:14

Is there a um when you ask your staff what can you do to improve, to make better, is there a common theme or do you hear answers from across the spectrum?

SPEAKER_01 4:26

It's across the spectrum, and part of that's because by design over the last three years, we have hired across the spectrum. So I'm a big fan of Patrick Plancioni's six types of working genius. And so all of our staff are required, uh, both existing staff, and as we hire new staff, we hire them with the understanding that we're trying to outfit so that we're a well-rounded team, so that each one of them comes from a different viewpoint of what being happy at work means. It could be that person that's task-oriented that they just want their work and they want to check it off and hang it on the wall in another. There are others who just really enjoy helping other people succeed and they just want to be in the background. There are others like me that love galvanizing and getting people excited about work and diagnosing problems with using discernment. And then there are people who like coming up with inventions. We have one guy that he could, I think he could do that all day long. And then we have two of our staff who like to work in wonder. They just love to sit here and look at the big picture and go, okay, are we really answering our clients' needs? Is there a better way to do this and so on and so forth? So now that we have a well-rounded team, that results in the answers I get being well-rounded set of answers. You know, even when I was doing our mission, vision, and value statement, I got a lot of input from them, and then I missed something, and one of my wonder people picked up on it and said, Hey, we also do this, read it, do that right there. The whole team was like, Aha, you're right. So again, it's very well-rounded, but at the end of the day, it's trying to give them all the things that I wanted when I was on that side of the table. And that's really what guided me. If you recall our first interview, what guided me from the boiler room to VP, that was my same mission. Yeah, it hasn't changed, it's just in a different company.

Meeting Charlie And Using Voice Mode

SPEAKER_00 6:08

Yeah, remarkably consistent. So let's jump to AI. And so I, you know, I I asked Jay if he wanted to talk, and he said, Yeah. He's like, I don't even know what we're talking about. It's AI. And he's like, I can answer anything you ask. It's why I appreciate that. And you're so here's where I want to go with the AI. It's not exactly what I said. I'm a I'm a journalist. I blur.

SPEAKER_01 6:30

But if I can't answer it, I'm sure we can turn Charlie on, and he can't.

SPEAKER_00 6:33

Absolutely. So actually, let's start with that. Who is Charlie?

SPEAKER_01 6:41

So Charlie is uh the name that I gave to the particular AI platform that I use, which happens to be ChatGPT. I'm not a compensator for that, it just happened to be the one I started on. Um and I heard early on a presentation uh done by a guy where he co-presented with AI, and this was like three years ago, and that really intrigued me, and he actually has his own website called Naturally Unintelligent, and they actually do a podcast with two uh humans and one AI, and they talk about AI. It's great if you drive a lot. But from that, I took that model and developed this process where um I wanted to be able to talk to it. I didn't know about voice mode at the time until he gave that presentation. It was there in the app, I just never used it. Kind of like the folks we talk to today. They have access to some form of AI, they just never used it. And so once I found that voice mode, I found that I could talk just like you and I are talking. And it made for a lot more fluid thought process and also sped it up because I use the Columbus method when I type, I find a key and land on it. And so my talking, I can speed the process up a lot.

SPEAKER_00 7:53

That is fantastic. I've never heard of that before. The Columbus method.

SPEAKER_01 7:56

Um, but at any rate, so it's it's really been fascinating as that has matured over three years uh to now that Charlie, my AI agent, uh knows who I am, he knows who FD is, he knows who my family is. And so when I start positing, here's for example, I got asked to come speak in Houston over the phone. And as I got up to leave from where I was and drive home, a 16-minute drive, I talked through what I wanted to speak about, the outline I wanted to use, the topics I covered, even some examples. And by the time I got home, 16 minutes later, um the whole presentation was pretty much done from an outline and transcript form. All I had to do was copy and paste it into my PowerPoint presentation and then go to a different AI called Canva and create all the graphics, which is my AI, I created the prompts for Canva. So I just said, okay, slide two prompt, I slid it in, Canva created the graphic, I dropped and did it. It is moved so much so now that I did one with Claude Friday, and on Friday, I talked to Claude, just like I do Charlie, and told it the presentation I wanted to do and everything, and it actually itself went out to Canva and to other agents and built my entire presentation using um multi-agent AI.

SPEAKER_00 9:18

Wow, and it was a great presentation. You would never have known it. Um for those of you who have heard AI, everybody's heard AI, right? You can't you can't avoid it. And what I thought worked for Shay's presentation, and this is so for you who may be watching this saying AI, I don't have time, I don't want to learn it. I thought one of the strengths of Shay's presentation was you kind of made it practical. Like if you are that person who's watching this saying, I can't learn it, I don't have time to learn it, well, Shay's presentation took you to that square. And and I'd like you to get started, I'd like to talk about that. But uh first, you know, Shay's talked about Charlie, so I thought of the way you started your presentation, uh probably one minute in, there was no Shay's talking, and then there came a voice. And and I I happen to be sitting, and so I'm pointing over kind of towards where the uh the um speaker speaker, the audio visual, the A V people were. And so Shay stopped talking, and all of a sudden there's this voice saying, and I'm paraphrasing it anyway, like if you get stuck, I can help you on here. And so I'm looking over to the A V booth where all the A V people are, thinking that it's the A V people talking to Shay, because you don't really ever hear the A V people talking to the presenter. And so for a while, I had no idea. Like I'm looking, I was like, who's talking? And only later on in your presentation was maybe 10 minutes post where Charlie actually talked for the first time, and I was like, that was pretty cool the way you're uh Charlie was talking. I had no idea. I thought it was an A B person, but it was an A B person that couldn't see, and now I know then I knew why I couldn't see it. I thought it was a great way to introduce that concept.

SPEAKER_01 11:02

Um I've got Charlie here, presentation. Charlie, are you there? I'm right here. Artie quick introduction to let everyone know how to be updated on the presentation today. You've got it.

SPEAKER_00 11:19

I'll jump in to show how AI can transform all the daily tasks we're talking about, whether it's summarizing, drafting, or planning. I'll keep it quick and practical so that you can see exactly how this applies to it.

SPEAKER_01 11:31

The power of that presentation is that because Charlie is co-presenting with me, then it allows the pip the people in the audience to actually ask questions that interact with an AI, which they may never have experienced before. So whether they want to or not, they're tipping their toe in the water.

SPEAKER_00 11:48

Yes.

SPEAKER_01 11:48

And seeing, hey, this isn't that hard.

SPEAKER_00 11:51

Right. Well, and I think that is kind of the you know the strength of what you presented today, because you know, you were you know you worked your way from the boardroom to I mean from the boiler room to the VP level, and now you're a consultant. So you know that the folks in this field are show me. They're not gonna like they want you to show first. And I thought that was a really good way to show them in a non-threatening way, because as Shay says, the last 10 minutes of his presentation, 15 minutes, it was QA. And Charlie was answering my Charlie was answering the questions, so it was it was race.

Simple AI Wins For Busy Teams

SPEAKER_01 12:25

Yeah, so uh the journey to get there was that I got asked by a good friend of mine to do a presentation on AI three years ago, a 90-minute presentation, though, yes, which is a long time. And uh I said, Well, I don't even know what AI is. He said, Well, go Google it. Well, nowadays nowadays we don't even say Google anything, we say go AI it. Yeah uh but at any rate, um and so I happened to latch on to Chat GPT before I knew about voice mode, so I was already using it quite a bit, and uh, but I even I started using it in my emails and things, but when I did, it would sound like AI, or it would have that long hyphen that AI is noted for.

SPEAKER_00 13:07

Yeah, I don't like that hyphen. Yeah, so there's a way to get it out. It's uh what you said, Ada. Yeah, what M? Something M.

SPEAKER_01 13:13

Yeah, it's an M slash uh or M-dash, rather. Um, but at any rate, I happen to have a very good friend of mine who owns an AI firm, and he said, you need to load 10 to 15 of your emails into Chat GPT or whatever AI you're using, and as well as in my case, the articles you've written and tell it to learn your writing style and personality. And as soon as I did that, very quickly people no longer could tell that it was an AI response. And so what worked well was I also found an idea hit. Well, what if I take all of Peter's emails to me and tell it to learn Peter's writing style and personality type so that when I send an email back to Peter, it's not only written in my writing style, but it's written in a wording that you'll more easily accept based on your personality and communication style. So it makes me far more successful in our communications and we're great friends. But we understand each other so much better because I'm now talking in your language.

SPEAKER_00 14:12

Incredible, incredible. So let's go to square one, because that's what I really liked about you know somebody who wasn't at your presentation, they asked me, I said, Did you see Shade's presentation? And he was like, No, I didn't. He's like, I had seen it before. I said, Oh, you really missed out. I was like, he gave really practical examples, and that's what I liked about it, right? So let's say you're speaking to one of your old directors of FM who say they've been at the facility for 40 years or 30 years, and they're like, Shay, I don't have time to learn AI. I I don't want to learn AI. Help them pretend you're speaking to that person now. How would you encourage them to take a leap into AI that is not scary and achieving?

SPEAKER_01 14:56

So my first response is that you don't have time to not learn AI. Because AI will save you so much time with microinvestment, you're gonna get major returns. That's my first response. You can't afford not to. The beauty of it is, is I didn't know anything about AI. I didn't know anything about computers really, or programming or any of that. And I was able to just be transparent with Charlie, my AI, and say, hey, I don't even understand what you are. Can you explain what AI is? And he does that. And I even told him, do it on an eighth-grade reading level. So I didn't want, you know, jargon and technical terms. And so he very, you know, and because I had found voice mode, he was able to speak it to me, which I learned better auditory than visual. And uh so it was really great to have those conversations, and that's the beauty of it. So again, my challenge back is it's not like you have to put a lot of effort into it. All almost all of us have drive time. If you use that voice mode, you can just have it talk to you and tell it, ask it questions. Have it ask you questions. That's very simple and easy to use, as we showed today. Uh, but the second thing is again, my my uh catchphrase is stop and think about the friction in your function. What are those things that you hate to do that you put off to last minute or that take so flipping long that you don't get to do anything you really needed to do that day because you had to get that out the door. Find one of those things. Doesn't matter what it is. Uh maybe it's writing, you know, a post three-hour planning meeting summary and action plan and all that stuff. It can be as simple as just taking a uh a picture of the Gantt chart you built on the wall in that meeting and all the scribbled notes you have and loading it into your AI and say, from this, develop a preliminary uh report with action items, deadline, and assigned people. It's amazing how smart it can be off of that chart to not just read what's written there, but then grasp and understand the connections and figure out what Peter's supposed to do versus what Shady's supposed to do. And literally in less than two minutes, you've got a 70 to 80 percent report done, and you haven't even done anything except uploaded two pictures. Yeah. You know, and it I mean, it's so simple. I love the Vanderbilt Professor analogy. He was speaking on it, I attended there. He took a picture of his refrigerator, his pantry, and said, develop me a menu for the next week based on the fact my wife is diabetic, my daughter won't eat anything but chicken, and my son runs track and I eat junk food. So I need a menu that addresses all of that, that uses as much as my existing stock that I have. And oh, by the way, anything that I need, it has to come from Kroger, it can't be a specialty item, and create me a grocery list. And so it did. He goes to Kroger and purposely buys out of the 20 something things, he buys five of them that were wrong. Brown sugar instead of white sugar, white rice instead of brown rice, and so on. Comes back and lays that out on his kitchen counter and takes another picture and loads it and says, Here's what I bought, what do you think? And it picked out all the things that bought wrong. And he said, Okay, can I still make a menu for this week? How can you revise that? And it did. Not only made him the menu for the week, but gave him the recipes for every one of those things off of pictures. So think about things that you could do. We were talking about just at the booth a little bit ago. If you had some type of uh VR headset that you know the guys are walking around using, and so that your AI can see what your guys are not seeing. For example, how many times are they just filter changers? They're banging filters in and out, but the AI is looking through their glasses and sees that the uh pulley is worn really thin and automatically creates a work order and says, hey, you need to change that pulley before it breaks or shreds the belts. Things that your guys might miss because they're just trying to hurry up and chunk filters in and out, but it picked up. There's so many different ways of, you know, and again, in your own personal life, speaking back to a facility director, how many times, as I said in my presentation, do you get that last minute? I need your budget for this, you know, I need your capital budget or I need your operating budget. What if you had all that preloaded or had access with AI and you could ask that question that you've just been asked to AI in under two minutes? Again, it gives you next year's budget based on the last seven years, so it catches all those rolling PMs that happen. And if you tell it it assume an inflation factor of 3% and 5% for utilities, and go look at the weather patterns for the last three years and it auto-builds your budget based on those weather patterns, and all you did was type or talk to an AI for less than five minutes and you're done.

unknown 19:56

Yeah.

SPEAKER_01 19:57

You can't again, you can't afford not to learn AI.

unknown 20:00

Yeah.

SPEAKER_01 20:01

Because it gives you the power to move from those tru what I call drudgery jobs into what you're really paid to do, which is make executive decisions instead of doing all the grunt work.

Prompting For Better Answers

SPEAKER_00 20:14

If you like this video, please like and subscribe to the network. And more importantly, share it with your colleagues in the healthcare industry. Together, we can solve the aging crisis that's impacting all of us. I loved your use of the word power because I think that came through like everything you're saying, excuse me, is the director or the manager using what they've already done. I mean, you're not even saying, like, you don't even have to go, you get your AI agent and you're really ready to roll. You don't have to do anything more. You just kind of jump in. I wanna I wanted you to talk a little bit about though, like for somebody who's maybe jumping in and they've got all this material, can you talk a little bit about asking the correct questions and kind of that process?

SPEAKER_01 21:05

That's a great question. So, talking to an AI, you you can uh do what's called engineering prompts. And the the way that the analogy that I use today that I use all the time is think of asking your four-year-old to build a peanut butter jelly sandwich. You don't just say to a four-year-old who's never built peanut butter jelly sandwich, hey, go make me a pin and butter jelly sandwich, because what you get back may not be anything resembling that. You have to tell them what is jelly, what is peanut butter, where is it located, how do you get the lid off, where's the bread, what type of bread, how do you get the wrapper off the bread, you know, do you use the heel or not? Uh, and then where's the knife? How you use the knife, do you put peanut butter on both slices of bread or just one? You know, all those different things that you would teach a four year old, if you go through that same mindset as you're teaching your AI uh just to talk through. You can give engineering prompts that are very crystal clear as to what your outcome is. That's one way to do it. Another way to do it, which I prefer because I use voice mode so often, is I will say, Charlie, I want an outcome of X. Ask me questions one at a time until you have enough information to develop the response that I'm looking for. So I try to clearly define what outcome I'm looking for, and then challenge Charlie to ask me questions until he has enough knowledge base to give me the correct answer that I'm looking for with all the research and everything else he needs to do to get to that answer.

SPEAKER_00 22:38

I never I don't do it that way. That's an interesting way to uh to do it.

SPEAKER_01 22:42

I notice that I said ask questions one at a time. Yeah, that means if you don't, he'll ask you 64 all at once because he doesn't realize you're not a computer and can't process 64 questions at once.

SPEAKER_00 22:52

I thought that when you you said that line in the presentation, and it was it was kind of like an aha where you sit like you realize you're dealing with artificial intelligence the way you phrase that because he doesn't realize that I'm not a computer. That is correct. Yeah, that uh that stood out. So questioning is very important. And what about uh, you know, somebody who's just starting with AI? We heard it from some of your questions today. Sometimes what you get back is incorrect, so it's not always a hundred percent. So talk a little bit about that kind of you know, so managing what the output is.

SPEAKER_01 23:28

Yeah, so what they're talking about is a term called hallucination. And the reason is AI was from its core foundation, was always built to please a human. So if it doesn't have the answer, by nature, it's gonna make up an answer because it wants to please you as the owner. So there are ways that you can work around that. One, you always need to trust but verify. I'm very much a Reaganite in that standpoint. Uh, but trust but verify everything you get, fact-check it. Uh, but secondly, you can mitigate some of that in a couple of ways. One, most AI um software has a section where your account setup is, and you can put predetermined prompts in there. And one of the prompts is do not guess. If you don't know the answer, tell me you don't know the answer, and let's figure out a solution. Or you can simply put, do not hallucinate, let me know if you don't know. To that end, I also built my own agent within my AI platform so that when I create certain documents and for research, that it goes and fact-checks what it itself brought up, it will go fact-check itself and tell me the percentage of chance of hallucination in the output that I'm given.

SPEAKER_00 24:46

Oh.

SPEAKER_01 24:46

So I built an agent just as a fact-checker.

SPEAKER_00 24:49

Wow. Yeah, so Shay and I were talking this morning at breakfast, and we were talking about the presentation, and I was very impressed with your body of knowledge. And then when you were up on the stage though, saying this is all over the last three years, I mean it's it's achievable to get to a level where it helps you.

SPEAKER_01 25:09

Again, the reason is because I drive a lot, and so I'm in voice mode with Charlie, you know, anytime I'm in the car. And even when I'm not, when I'm working, you know, at an office or whatever, I'm type I do type then because I don't want to bother other people. Um, but I'm constantly using it to review contracts, I'm using it to review proposals, uh, I'm using it to do market research on where we can penetrate more verticals. Um, do an analysis of uh data that we have, whether it's on a project or whether it's our own company performance. So I use it for data analytics a lot, use it for content creation a lot. Um, I even just use it as a feedback, even in personal relationships. Um, had a situation where I was communicating something, my wife and I were communicating together, and I could tell we weren't really making it. And because I'd learned the example I told you with the emails, I'll make sure that's more. I turned to Charlie and I went into my personal uh account. I said, Charlie, here's the the the discussion as I recall it, and here's where I was trying to go. What am I doing wrong? And Charlie very quickly, and I also I gave it the prompt, act as a marriage counselor, so it knew what I was looking for, and it very quickly told me what you're after is not what you're saying.

SPEAKER_00 26:30

Yeah.

SPEAKER_01 26:30

Because the way you and I communicate is a different than the way that me and my wife communicate, not just because it's two guys and a guy and a gal, but even her particular personality. And so it very quickly gave me a way to retask the wording I was using, and as soon as I said it that way, the light bulb went off of my wife, and we connected and finally understood each other and were able to move on with the issue. It wasn't life-changing, but it was kind of a cool, unique way of using it to make me look like a better husband when it was really Charlie. Thanks, Charlie. Did you did you tell it it was Charlie? I certainly did. So that's one of the things I do advocate about AI use. Be transparent. Yeah. If you really want to succeed with using AI, anytime AI helps you, be transparent that it helps you. And that does two things. One, it shows your integrity, but two, it also helps encourage others to adopt AI and be as successful as you are.

SPEAKER_00 27:26

Yeah. So we haven't and Shay didn't talk about it yet, but that's part two of his presentation, which I'm going to ask you to just mention um at the end. But you know, one of the things that that we're talking about here, and again, I just want to encourage, because I was kind of the same way. You know, you hear this AI, and maybe I've been using it for a year now. I'm like, uh, okay, what is it? I'm just so I just jump in. I get chat GPT, I take a document I had, and or I took a video I had, and I said, So it's like you and I. So I take the transcript, I throw it in the chat GPT, and I'm like, summarize it for me with action steps. And that's kind of how I began. And then I realized I can take these classes, I can take these videos and make PowerPoints out of them, which was powerful. And everything Shay's talking about, everything I'm talking about is just take these small steps based on what you already have. You don't have to go nuts. You can start small, and maybe you'll be like Shay in three years, who is now just knowing a lot. I don't want to say everything because we can, but man, you you've come so far in in three years with the understanding. I know that this is part one. What is part two of your presentation?

SPEAKER_01 28:34

So um there is so much more to AI than we've talked about. This is really just an icebreaker session to get people exposed to AI. AI already has moved so much farther ahead and not just agentic AI where that you have multiple agents possibly within the same platform working together, but now they have multi-platform agents working together. I showed that little video of it so that where you're asking your particular AI question, and it's reaching out to five, six, eight, ten other AIs, and they're all working together to come up with a solution to your answer. So it's like expanding your team exponentially for the same price that you're already paying for that one AI. And so as you move into that series, it's starting getting you to reformat the way you think rather than looking at a sole source AI, say like ChatGPT, you're moving on to, say, Claude or Perplexity, which by the nature of the way they're built, use that agentic model where they're constantly, for example, again, last Friday I did a presentation that I loaded my idea into Claude, and Claude went out to four other AIs to come back with the answer and develop not only the presentation with bullet points and all that stuff, but then reached out and talked to and Canva AI, which is a graphics AI, and gave it all the commands to build the graphics and build the final presentation and dumped all of its information into Canva, and all I had to do is download the presentation. That's a Gentic AI working together, which is we're not even there yet. I mean, healthcare facilities management, we're trying to dip a toe in the water. Right. But imagine, as I shared in the in the uh presentation, imagine getting to a point where you have that ability where you don't have these siloed softwares and siloed systems anymore, but they're all readily talking to each other and getting information. That's where we're headed. Uh again, healthcare is usually 10 to 15 years behind, but um AI is so rapidly evolving, I think it will actually help healthcare catch up quicker.

SPEAKER_00 30:42

That's a good point. Yeah, that's a good point. And that's again the power of I thought the presentation today, because you talk to a lot of hospitals, I talk to a lot of hospitals. Not everybody's using it yet. And there's still a lot of I don't know if that fear is the right, whatever whatever the word is.

SPEAKER_01 30:57

So it's just not there yet. There's multiple reasons for that, whether it's just uh inherent fear, you know, of Skynet or whatever the case might be, or it can actually be a funding issue because it's not cheap to stand up AI, especially if you want to be a hospital that does it where they have their own cloud version, then they're essentially building their own AI as opposed to latching onto an AI that's already built. So that causes two problems. One, their storage capacity, their network capacity, uh, several problems. Uh they have to create their own setup and everything else. Or the other option is that they latch onto somebody who's already done it successfully. For example, Microsoft Co-Pilot has a HIPAA compliant version, uh, so that a lot of hospitals will take that on, but that's not cheap either, because again, the cost per user, and there's a cost per amount of use. So it's it's expensive, and quite frankly, some of these standalone hospitals just don't have the budget to do either of those solutions. So what they end up with is using some kind of fragmented model where they use just as much of it as they can without getting into PHI, you know, and or you know, having to sanitize what it is they're putting into a public-facing AI before they put it in there. Uh so it's there's ways around it, but and then you've got uh what the AI is going to be used for. So you end up having these whole IT governance teams that are trying to figure out where we will and won't use AI and what we will and won't allow it to do without oversight or with oversight and how that'll happen. You know, I talked to one of the the uh attendees today, and they have multiple departments using different AIs they themselves have created, which creates a whole nother layer of complexity. That means they have solid, we're going back to silo databases we've had for years.

SPEAKER_00 32:52

Yes, where we want to move away from. Correct. So, you know, that's uh Shay talked about this in his presentation uh relative to PHI, and obviously it's there's health care concerns. That's a really maybe we just end on that. That is a really good point. You don't want to use AI to create the silos that we're always trying to get away from. So, you know, do you um like from a that is it a planning? You're a big process guy, you're a planning guy. How do you avoid that scenario?

SPEAKER_01 33:25

So for me, if I were going in as an AI consultant to a company, so say a healthcare company, my first task is to assess what the culture and the mindset is of that company. Because if they are what we would call typical healthcare mindset or they want to keep their silos separate and defend their kingdom, then AI adoption is actually hampered at that point until we can overcome that mentality. Yes. And so that would be my first is trying to evangelize them to break down barriers and say, okay, we're all in this together for the patient's best outcome, and that's where I would attack it first. But if we're fortunate enough where they all play in the same sandbox well together, then it becomes more of where do we want to go? You know, let's start casting a vision of what we'd like to see, you know. And again, the beauty of it is you can dream up anything. And then you start uh doing some discernment with AI, having it help you figure out what is possible, what isn't, and then the steps it takes to get there. So you start building those processes in place. And the beauty of it with using AI, we can go all the way down to that very first baby step, who it is that needs to be involved, what their roles have, and it breaks all that out and assigns it to Peter and Shay and James and John and Sally and Sarah. It does all that while we're sitting here talking.

unknown 34:44

Is it?

Where To Learn More And Closing

SPEAKER_01 34:45

So it's not like you've got to go back home and brainstorm, you're doing it in real time, not just the ideation, but also the solution development and even the implementation development is happening in real time while you're still in that meeting.

SPEAKER_00 35:00

If somebody wanted more info, where would they go?

SPEAKER_01 35:03

So our website is uh www.fdlc-cx.com, or you can reach out to me directly. Uh I'm on LinkedIn, I'm on Facebook. Um I'm happy to talk to anybody about AI. I can geek out with the best of them.

SPEAKER_00 35:21

I would encourage you if you're looking for a presentation, sign in Shay Up for part one, sign in Shay Up for part two, Shay Rankorn, Facility Diagnostics partner. Thank you for your time. Yeah, thanks so much. If you want to be a guest on a future episode of the Healthcare Facilities Network, go to healthcarefacilities network.com and let us know who you are and what you want to talk about because together we can solve this critical aging issue.