Subscribe
Share
Share
Embed
#74 Tribute Home Care has no engineers, so Andrew Dobson stepped up and is helping build AI tools in-house that collectively save 300+ hours per month. He explains how a scheduling dashboard evolved into an availability analyzer that helps staff new cases faster at 90–95% caregiver utilization. Tribute also used a Viv-built voice agent to fill 75% of its open weekend visits in an early test. Andrew shares how they condensed a five-system lead process into one workflow and reduced two days of financial reporting to a button press. He offers practical advice for finding the right AI projects, developing an internal champion, validating outputs, and keeping humans involved wherever judgment and relationships matter.
Sponsors:
- Phoebe—AI scheduling teammate that fills 75% of shifts in under 15 minutes
- Reel Home Care Consulting—my recommended consultant for startups focused on differentiation and delivering a hospitality-like experience
Creators and Guests
What is Home Care Strategy Lab?
Is there a single right way to run a home care agency? We sure don’t think so. That’s why we’re interviewing home care leaders across the industry and asking them tough questions about the strategies, operations, and decisions behind their success. Join host Miriam Allred, veteran home care podcaster known for Home Care U and Vision: The Home Care Leaders’ Podcast, as she puts high-growth home care agencies under the microscope to see what works, what doesn’t, and why. Get ready to listen, learn, and build the winning formula for your own success. In the Home Care Strategy Lab, you are the scientist.
Miriam Allred (00:01)
Hey everyone, welcome to the Home Care Strategy Lab. I'm your host, Miriam Alred. It's great to be back with you this week. Today in the lab, I am joined by Andrew Dobson, the Director of Technology Innovation at Tribute Home Care, based in Massachusetts. Andrew, welcome to the show. It's great to be with you. How
Andrew (00:16)
Hi Miriam. Great to be here. Great to be with you.
Miriam Allred (00:20)
How are you doing? How are you feeling today and this week?
Andrew (00:23)
I'm doing great. It's a nice labor day, a little three day weekend, so happy about that and just happy to be here with you today. Yes. I am an avid listener, yes, I've
Miriam Allred (00:30)
Yeah, likewise. Your first podcast recording, but you're no stranger to the show. You're a I I don't know, avid listener?
Andrew (00:37)
I've heard you quite a lot over the years, so it's exciting to be here.
Miriam Allred (00:39)
Yeah, so excited to excited to put you in the hot seat. yeah, just a little bit of context for people listening. I'm gonna be interviewing and kind of co-presenting with Andrew at the HCAOA in a few weeks. And so we've been talking about the concepts and putting together the decks and kind of working through our flows. And you've shared a lot of really great things with me. And I was like, I think I wanna interview you you on the show because there's a lot of people that listen that are probably not gonna be at HCAOA and they are gonna miss out because our session is going to be top tier but hopefully this will kind of make up some of the difference. So excited to jump in and talk about AI. I wanted to ask you right out of the gate, like are you tired of talking about AI and thinking about AI or no?
Andrew (01:20)
I I feel like I don't talk about it enough to be honest with you. LikeI work through it all day long and I build these projects and I've got all this fun and you know exciting information to show everybody about how they can make their jobs easier. And I just, you know, I love talking about it. I love showing people these tools, I love transforming their jobs and making things easier. So no, I'm not sick of it yet. We'll we'll see what happens though.
Miriam Allred (01:45)
Do you spend a lot of time online though? Do you feel like your social feed or your news feed, it's like all you ever see and hear about is AI or no?
Andrew (01:51)
yes. I mean when I'm going through my reels on Instagram now, I feel like every time I'm clicking up it's another thing about Claude or, you know, somebody's story with AI or how AI is taking over the world or or whatever. Yeah, yeah.
Miriam Allred (02:03)
Okay, 'cause that's the fatigue I'm talking about. Like I just feel ike it's all anyone is talking about online. And so I'm getting this like kinda like online AI fatigue. but like you, it's like, no, it's still exciting and interesting and I'm using it and wanna like get better at it, but just like online, I'm just like, Okay, let's talk about something else.
Andrew (02:20)
Well, I think you see it so much in business discussion nowadays. Every time you turn on the news, AI is taking over, it's taken this many jobs. And I think for so long we heard about it. You know, ChatGPT came out and there were, you know, opportunities for us to prompt and make our emails easier, but
I think, you know, earlier this year when Claude had come out and released some of their new features around building and, you know, vibe coding and all that, it just really changed the landscape for me personally and I know for a lot of other people to really start getting into it and using it to kind of, build tools to make life a little bit easier. So yeah, I'm still stuck on that momentum, but we'll we'll see what happens.
Miriam Allred (02:59)
Yeah,
do you feel
like your journey though is parallel to a lot of people of like, Yeah, we were all talking about it and hearing about it like a couple of years ago and it's just gotten progressively better, but like a lot of acceleration the last like six months or twelve months, what would you
Andrew (03:11)
yes.
Miriam Allred (03:12)
say?
Andrew (03:12)
One hundred percent. our CEO has, you know, really been in it since the beginning, John Sneath, and we've been very lucky enough to have him kind of promote it and, give all of the office associates Chat GPT and Claude and everything like that for us to be able to discover it. So yeah, I think that things have blown up and it's been nice to be able to kind of experiment for as long as I have and finally be at that spot where it's like, okay, I feel comfortable with this now. I can start really, making those changes.
So
Miriam Allred (03:42)
But none
of you are engineers. You're not an engineer background. John's
Andrew (03:44)
No.
Miriam Allred (03:45)
not. Do you have any engineers at Tribute?
Andrew (03:47)
We have
none at all.
Miriam Allred (03:50)
Which is just like a healthy reminder of like what we're about to talk about and the expertise that you're gonna share is like you're
Andrew (03:54)
Yes.
Miriam Allred (03:54)
not an engineer, neither is John neither is anybody. Most of you are like home care people.
Andrew (03:59)
We are home care people. I mean, I started working when I was twenty years old as a caregiver and here I am now. So
Miriam Allred (04:06)
Yeah. Tell
us more about that. I I know I'm just jumping in right here, but yeah, introduce yourself a little bit more for the people that don't know you. T I didn't even know you were
Andrew (04:09)
Sure, yes.
Miriam Allred (04:13)
a caregiver at twenty. That's wild.
Andrew (04:15)
So
you know, at 18 I wanted to work and go to school at the same time, and my friend was like, Hey, I'm working this job with adults with intellectual disabilities, and it's really fun, and we do all this cool stuff where we go out each day and take people out. And so he gave me the referral and I applied and got, you know, hired, and I did that for a few years through college, and it was eye-opening. Like I I you know, they talk a lot about, people with intellectual disabilities kind of like living in the shadows. You never really see
see that until you really step into it. And I really loved it. You know, building relationships and being around people. So that kind of, influenced me to get a degree in, psychology. And from there I worked in schools and was a director at a couple group homes.
And then became a stay-at-home dad for a few years where I almost got to be a caregiver for my own children. and then I found a job for Tribute Home Care as an after hours manager. And I was like, this is kind of interesting. I could potentially work at night and still be with my kids during the day. And so applied for at Tribute and did that for a couple of years actually. loved it. And then they had an opening for scheduling, so I was like, I might I might.
actually
doing scheduling and you know they they had a lot of flexibility for me at the time and you know just I've kind of worked myself up from there, became scheduling director of operations and then here I am now director of technology innovation so it's it's been a ride and I feel like having all of the
kind of understanding of what it's like to be on the ground up through operations has really kind of allowed me to be in this place where I understand, what it's like for everybody. And it's just not a lot of people get that experience. And I feel very grateful for it. So
Miriam Allred (06:04)
How long have you been at Tribute?
Andrew (06:07)
I am going on my sixth next month actually. So yeah.
Miriam Allred (06:09)
Okay. And how long
had they been around prior to you joining?
Andrew (06:14)
Tribute. don't quote me on this. Sorry, John. 2012, I believe we started with John in Baltimore, and from there a couple years after we grew into Massachusetts and from Massachusetts we opened up an operation in Chicago, I believe around tw 2022. and we're going strong. Things are looking very great for us. Yeah. Mm-hmm.
Miriam Allred (06:37)
Okay. Okay. So they had been around for six to eight years and then you've been
there now for six or s six or so years and they've grown a lot. Would you say you've experienced kind of like the bulk of their growth? Because yeah, like you just mentioned, they've gone into other markets and things. Have you experienced most of that growth?
Andrew (06:53)
yes, I've been here throughout, you know, a lot of the growth that we've experienced and it's been quite the ride and you know it's the momentum is still there. So it's it's quite fascinating to kind of just be experiencing that. Yeah.
Miriam Allred (07:06)
Yeah. And it sounds like organically
over the last, I don't know, six to twelve months, you've kind of become the AI guy. Granted, it sounds like
Andrew (07:13)
Yeah.
Miriam Allred (07:13)
John like gave everyone the technology and was kind of like an early adopter and like exposed everybody to it, but you have kind of become like the point person. Why is that and how did that come about?
Andrew (07:24)
You know, I
When I was director of scheduling operations, there was a lot going on. You know, we had I was overseeing the after hours team, the admin team, scheduling and you know, pulling all the reports day to day, stepping into scheduling when I needed to do that, after hours, and it was just a lot. And I was like, this AI stuff seems like it might offer me a solution. I know we're using it to kind of write emails right now and just give advice about things, but you know, what if I could start actually building
Tools with it. So, I I think what I had done was I started with reporting. You know, we had our weekly KSFs that we have to complete, and it was a lot of manual uploads going through VIV and pulling the schedule for the week and our gross margin report and things like that. And I was like, I wonder if I just kind of like take all of this in bulk and just you know run it through ChatGPT at the time to see if it can kind of replicate what I'm doing. And so it was a lot of
trial and error of like you know putting it through seeing what it generates kind of you know understanding like you know or I'd have to do a lot of validation there and you know finally got to a place where I had my scheduling KSF like fully automated where I could just kind of put in you know the documents and it was done and I'm like that just took away five hours of work I wonder what else I can do so it really was this past year where I just started thinking of every single process that we're doing you know that or that I
Was doing and people on my team at the time, and finding ways where we could take that process and kind of have AI support us in kind of you know semi-automating, not quite full automation, where we have agents running everything, but just like simple click of a button. that report's done. that dashboard is showing us what we need for caregiver utilization this week, or something like that. So I think from there and just kind of you know having that motivation,
We realize as a company, well, this is happening fast. And you know, we really need somebody that's going to be kind of like
pulling all of our, you know, KSFs that we have, putting them into dashboards, and looking at some of the other processes that we do outside of scheduling. And I was like, I'll do it, you know, sign me
Miriam Allred (09:38)
Mm-hmm.
Andrew (09:39)
up. And so you know, I probably not the best example here, but five o'clock would happen and I would be on, going through learning how to use Claude, lot Claude code. And I was in a position where it was up until midnight sometimes because I was so fascinated by this and just
creating all this stuff that you know talked to John and other people that you know this was you know really the path that I wanted to be on and they were in full support of it because you know it was it really allowing us to be in a position where we can start to scale without having to hire from within which I think is you know the big drive for a lot of companies right now. And you know a couple other people on our team decided you know why don't we start really looking at how we can transform as a company throughout these next couple of years.
know not only with AI, AI definitely being a huge part of that, but also as a team just looking at how we can become more efficient. And so we're currently in that process right now and I'm driving a lot of these tools that are helping us to get through that transformation. So yeah, that's kind of the you know the story in a nutshell.
Miriam Allred (10:45)
Yeah, that's great.
I hope everyone d that just heard that is like thinking, who on my team can become this? Like you had kind of like the knack and the skills and also the background, but then also kind of like volunteered of like, hey, I think this is interesting. Like, I want to spend my time here. There sh hopefully is or should be someone on every team that like has like kind of a knack and a passion for this.
Andrew (11:05)
Mm-hmm.
Miriam Allred (11:06)
And I would be like encouraging those people all day because this is part of why I wanted to interview of you, of like
you have become this like AI guru at Tribute. Every company in my mind kind of like needs one of those. Like granted, there's all these great AI point solutions and I'm not knocking on any of them, but it's like you need an internal like advocate and champion and someone that with a knack for this if you want to like adopt it quickly and
Andrew (11:32)
Mm-hmm.
Miriam Allred (11:33)
embed it like in the company. Like you need someone internally, bar none, even if you go outside and find other tools. So that's kind of like insight number one.
Andrew (11:41)
Yeah.
Miriam Allred (11:42)
Going
back a little bit, so think back to when you built that first kind of like reporting function for scheduling.
Andrew (11:48)
Mm-hmm.
Miriam Allred (11:48)
That was in Chat GPT. How long this is a while ago, but like how long did it take you to build that? Like how many hours, how many days, how many weeks? Like conceptually, how long did that take?
Andrew (11:58)
I mean that to get to a place where you know we were actually using, I would say about two months of just kind of like going in there and trying to replicate that. But that, you know, that was in the early chat GPT stages where it would give you something and you were like, no, this isn't right. so not the best example in terms of like how much time it would take now, but I think you know, for me to start learning how to do something like this, I really had to just throw myself in there and kind of like, see in depth.
what it would take to actually build what we need each week. And that that was our caregiver utilization dashboard that, you know, a really allows us to determine how much workforce we have to be able to take on new leads week to week. And it was a pretty important document. So I just wanted to make sure that, what we're seeing there, is actually the real data and not, having, gone through hallucinations and AI and things like that. So it it had taken a lot of time in the beginning, but
But from there it really led me to be in a place where like, okay, now I know what to prompt and how to say this and do do this to be able to, you know, make it a little bit faster for some of the other things that we do. Yes.
Miriam Allred (13:01)
And how to keep pounding on it. I think there's this misconception
of like, you can build these things really quickly and easily and then they're functioning. Like two months, like that's a good reality check. Granted, this was a while ago, but it's like it takes
Andrew (13:11)
Mm-hmm.
Miriam Allred (13:12)
a lot of time and a lot of refinement and a lot of just like pounding to get something
Andrew (13:16)
Yes, yes.
Miriam Allred (13:17)
even remotely good, let alone something that you can like roll out internally. So you started with if I'm understanding right, like you started I mean, your vein was like scheduling, but then you were like, Okay, reporting.
was an easy way to start. What were some of the other first things, maybe even just after that initial one? Where did you you kind of like did an audit of all of like the teams and processes? What were some of the other first things that you found were like easy to plug
Andrew (13:46)
Mm-hmm.
Miriam Allred (13:46)
AI into?
Andrew (13:47)
Yeah, so I mean a lot of it was that scheduling functions, you know. looking at availability, looking at, I think all of us, that deal with this tight utilization with caregivers, where, you know, Tribute's I I think you spoke to Jen Kulig while ago. We only hire one percent of our actual workforce. So, you know, we're dealing with like ninety to ninety-five percent utilization of our entire caregiver
workforce right now and that has been challenging. So, really going nitty-gritty into the availability and getting that information. so I built an entire scheduling suite from there. And that really was my kind of like prize thing where I was like, this is great. I feel amazing. it's really helpful. All the schedulers are using it each day and it's saving a ton of time. And we really got to start to look at even scheduling as a whole. Like, okay, now we can really start kind of, you know,
Centralizing them in a way where, we have two scheduling partners in Massachusetts, two in Maryland. And we're like, what if we, because we have this availability information so exact now, you know, take that scheduling team and split two of them into only managing new business leads to determine, how we can build a care team for them, and then you know, just have somebody dealing with the call-outs day to day. You know, if we split those functions up, with these tools, it'll really allow us to move a little bit faster and also deal with some of the redundancy issues that we have with the team.
so got in a really good place there and then started working with our finance team and others to look at some more of those weekly KSF reports. So, a big one that we do each week is our caregiver hiring. and that was really our first kind of like okay, we're gonna automate this kind of data sheet that we use each week that takes a ton of time to upload and just make it look a little prettier and you know, you know, have it be something that everybody can use. So yeah, we got
Got
into the caregiver hiring pipeline and really broke it down like this is how many applications we have each week, and this is the overall percentage of the caregivers that are applying that we actually hire, and seeing that visually and looking at that each week was you know tremendous value to a lot of people. and then we went into more kind of like gross margin and automating our financial statements. It really just blew up from there, and like everything that we thought that we could automate kind of just went into this full kind of like, you know, we can do that, great, let's get
Yeah, working on it.
Miriam Allred (16:05)
Can we talk
a little bit more about the scheduling tool? I think that like piques everyone's interest because everyone's
Andrew (16:09)
Sure.
Miriam Allred (16:10)
trying to like solve scheduling as best as they can. And that's why a lot of these like AI technology companies are like trying to solve that first because that's where a lot of the headaches are. you're
Andrew (16:18)
Yeah.
Miriam Allred (16:19)
citing Viv. I want people to understand that's your scheduling software. It's called Viv. sounds like you
Andrew (16:22)
Yes. Viv technologies, yep.
Miriam Allred (16:24)
built kind of like a a tool outside of Viv that was helping with scheduling initially.
Did you realize like out of the gate that you had to like integrate those two things or did you build something like that was a complete standalone outside of Viv and then you were manually like pumping information back and forth? Or do you have to connect your scheduling software like immediately for anything to be effective?
Andrew (16:47)
No, we realized at the beginning that it was actually much easier and allowed us to validate a little bit easier if we actually downloaded the reports and then uploaded them directly in the tools that we were providing. That way we could see the information, we could see what it was giving us as an output. So, no, you do not at first. after a while, yes, we did connect via API and worked with Viv to get that done, and they're constantly giving us more API connections to allow us to automate some things, but no, it a lot of
It was just manual uploads of reports, which still saved
Miriam Allred (17:19)
Okay.
Andrew (17:20)
us a ton of time. But to go into the scheduling tools, yeah, I can you know talk a little bit about that if you want. That
Miriam Allred (17:26)
Yeah, you mentioned like utilization and availability. Like these are all the buzzwords.
Like what ex what data were you pulling out and then what were you doing with it? Yeah, in the in the software that you built.
Andrew (17:36)
So I think at Tribute, our I think our biggest struggle is availability week to week. We have a task force where a ton of department leaders get together in one room and we look at where we're at in specific markets or just overall for you know long hours and short hour caregivers and things like that.
To really, get hiring together, to think outside the box and be creative. And so that really opened up the door for me to be like, okay, you know, I can take a lot of that information that we're gathering and ideas that we have from those weekly meetings to put into some tools that I'm making. so one of the big problems that we have, yes, is this caregiver utilization. It's not the best number to look at because we've really found that, okay, yeah, we can see that we have a 90% utilization of long-hour day caregivers.
The Massachusetts area, but we're not looking at restrictions. we've got 10 caregivers that might be on leave for the week or that you know can't that have certain lifting restrictions or things like that that Viv really just isn't putting in a central place, and I don't think a lot of other companies are either. so part of what I thought was okay, let's capture that information about some of these restrictions and really put it in a place where you can filter out for some of that. Like, does this care?
Giver work with dogs, does this caregiver work with cats? And so from there it brought us into this tool we call it, we call it the availability analyzer. And I was like, okay, let's put in an address or have it, you know, build this tool where you can, you know, you can put in an address, put what the new client is looking for. Let's say they're looking for you know someone that can work with dogs 247, and they require somebody with hoyer experience, and I was able to gather all
that information via a report from Viv and put upload it here and it would spit out all of those caregivers via their address and things like that and really allow us to see okay these are actually the caregivers that are available for that type of client in this area.
And it really went down into kind of like the nitty-gritty of you know what we were looking for. So it allows to go kind of like get into the individual kind of business situations where you have a new lead, we need somebody, here's the availability for that. But you could really also filter it out in terms of like getting a visual of what that true availability is, if that makes sense. which really is kind of like the core thing that we're looking for when we're saying at the beginning of the week, okay, what's our availability to take on 324?
Four sevens at Boston because that's what's going to allow us to make goal. And this really allows us to get that answer, and the scheduling team can respond in a much quicker way than they could before. Because I think before you really had to go through a list and literally look at every single caregiver, decide if they're available, when they're available, you know, all of those little things that I talked about. And it just took too long. I know we want a quick turnaround to be able to give a client an answer because we know they're going down that list to the next company each time they call. So
So that tool was extremely valuable and you know one of the first kind of you know takeaways that we got from the scheduling suite. and then we s him
Miriam Allred (20:47)
Was your was your caregiver data
good enough to be successful with that at the beginning? Or did you find quickly that you had gaps in your caregiver data?
Andrew (20:59)
So it allowed us to kind of tighten it up a little bit. So, I think a lot of ERMs like Viv and other companies, they
They allow you to kind of input that information, but you know, when you're trying to actually fill a visit or do things like that, it doesn't all spit out at you. So it was there, but we just kind of had to organize it in a different way to really make this tool a little bit more successful. I do think, and we can all talk about this for hours, that caregiver availability in this field really changes week to week. It's such a fluid thing. So we're only getting the information that we're being told from the caregiver week to week. So that
does change and I think we're constantly thinking about processes that you know we can really have people own that availability a little bit more to check in with caregivers. But for the most part, yes, we got to a place where, you know, we got that information to be pretty accurate. Yeah.
Miriam Allred (21:54)
Okay. Yeah, I would agree
with that. It's like all this information for the most part is in existence. It's just like these software companies haven't figured out how to surface all of that information, plug it all together, and then like formulaically, like, like give you good
Andrew (22:09)
Yes, yes.
Miriam Allred (22:09)
outputs. They're trying. Every company has a carefinder, a care match, a something, like a name for it, but no company
Andrew (22:15)
Yeah. Uh-huh.
Miriam Allred (22:17)
thinks, no agency is using it because it's not good enough. And so everyone's building these like proprietary workflows.
to put it all together. But I was just gonna ask that the availability piece, that is still on the caregiver to tell you and it's ever changing. Have you unlocked any processes or notifications or messages
Andrew (22:36)
Mm-hmm.
Miriam Allred (22:37)
to try and stay on top of that?
Andrew (22:40)
So we had to do kind of a systems change around this availability piece where it's always going back and forth between our caregiver excellence team, you know, they're the managers of our caregivers and check in with them, you know, on a you know, regular cadence, and then the scheduling team. I personally feel like because scheduling is constantly calling up caregivers to ask them to take on visits, that it made more sense to have scheduling own that. And I mean, the only problem we're really having is that, you know, scheduling is, you know, so busy we
week to week they're going down that list calling 50 people and then for them to have to record that information about that availability change it's not always going to be perfect and I do think it's better to have them doing it than people who are not checking with them as much as scheduling does. So
we've we've tried to make it easier and you know, we can talk a little bit more about some of the tools that we've built for that a little bit later, but I do really think there is something to, getting those phone calls recording eventually and you know, from those calls, having agents kind of pick through that availability, change notification or whatever that comes up. And you know, eventually I'd like to be in a place where it's just happening automatically, but we'll get there eventually.
Miriam Allred (23:51)
Yeah, where the schedulers aren't writing
down or documenting anything. It's all it's all
Andrew (23:54)
Yes.
Miriam Allred (23:55)
automated, transcribed, but then like plugged into Viv
Andrew (23:58)
Mm-hmm.
Miriam Allred (23:59)
or your software like automatically. Yeah, we'll get there. One one step at a time.
Andrew (24:01)
Yes, yes, yes. We're close to it. We'll get there. Yeah.
Miriam Allred (24:06)
something that you shared with me is like this like kind of pressure test concept for evaluating projects and workflows that are like ready for AI. I think a lot of agencies are in that stage of like.
Auditing their organization and figuring out where to prioritize AI because I think a lot of people are getting this like analysis paralysis of like, my gosh, like AI can be plugged in everywhere, but we don't have time. And like that just like hangs people up. But how do you, how have you decided like, okay, what's ready for AI? Where's the best use of our time so that we're not getting like hung up in different places?
Andrew (24:42)
It's so easy, I tell you, now that I'm in this new role, to just like jump from one place to the next right now, just because there are so many different opportunities available. But we've done a really good job in developing this transformation plan for the company. and and we kind of said to everybody, okay, what can we do right now, AI or not, to be able to grow without having to hire people?
And that's really caused us to have to look at those day-to-day processes. So we've broken it out by department, and each department leader developed a goal or you know, several goals to allow us to get there. And through that, kind of we've been able to kind of highlight, you know, specific tools that can be created to help them, you know, meet that goal. And for example, our caregiver excellence team, they go through this process of, you know, we call it our our rise
meeting where, you know, we're checking with caregivers on a regular cadence. you know, looking at them and saying, okay, you have all of these processes where you're collecting information about performance, you're collecting information about kind of celebrations, things like that. And I know there are a lot of tools that are being created for that now, but we have it all. And you're spending a ton of time, you know, going to that.
dashboard and then you know getting information here, you know, just to prepare for one caregiver meeting when they have 70 caregivers on their roster, you know, what can we do to really automate that and get that in one place? So things like that, where, we're focusing on each department, looking at their day-to-day processes, finding the solution to make it easier and allow us to meet that goal of not having to hire to, scale, that's really allowed us to be in a kind of a structured place to be able to get that done. So
Miriam Allred (26:29)
And what is the
lowest hanging fruit? Like, do you ask them like literally like where are you still writing stuff down? Where are you typing stuff? And then where are you passing information? Like to me, like how do how have you found like the low hanging fruit? What question I cause I'm I'm imagining I guess all these departments do kind of this self audit and then you guys come together and then you are
Andrew (26:48)
Mm-hmm.
Miriam Allred (26:48)
helping them like read between the lines and decide where this is effective. Like so what what follow up questions do you ask them to identify the lowest hanging fruit?
Andrew (26:57)
so
I've developed kind of like a weekly structure where I'm meeting with each department leader. And then we also have designated an AI key user for each department. So really just meeting with them and you know, okay, here's the tools we've created. What else are you doing day to day? And then meet yeah, exactly. Just extracting, like, well, you know, I have to type this email to this long-term care insurance provider to make sure that they get all the information they need week to week, you know, just that very generalized version of that. But you know, just hearing them say that and saying,
well, you know, if that was automated or if we could find a way to, you know, take that email and have it automatically created by Claude, would that save you some time? And they're like, that would save me hours. So just, you know, hearing that piece, week by week with all the different leaders and people that are in operations allows me to really see it and start thinking about ways that we can do that, if that makes sense. Yeah.
Miriam Allred (27:51)
Yeah, you guys are a lot
larger than a lot of agencies. And so you have kind of like the infrastructure and the team to be able to do this. Put yourself in the shoes of like a much smaller operation that has like a CEO, maybe like a you know, a branch director or director of operations, and then kind of like, you know, one or two people in each of these seats. How do they do this at a smaller scale? The director of operations or the owner? Like who should be helping with this like auditing process?
Andrew (28:16)
I mean, I definitely think someone who is in operations, so it has to be, you know, director of operations or, you know, if if you're smaller than you don't even have a director of operations, then you know, you get you have to have Claude or Chat GPT or something just at the corner of your screen. And every time something comes up that takes a little bit of manual, labor for you to have to pull a report or, collect information or send an email, just write it out. Put it on a even just put it on a list, a notepad, and then feed that into, you know, your AI, you know.
platform and just say, what can I do about this? How can I automate it? What can make this better? And then just work through it. And I think that's really where you have to be right now in order to kind of get yourself to that point.
Miriam Allred (28:57)
You make it kind of sound like and this is just the way that I'm like interpreting this, that like everyone at Tribute is like pretty tech forward. Is that actually the case? Or have you had a good amount of like your executives, your operators out in these different markets that aren't super tech forward? And has it been a challenge to get them to embrace this?
Andrew (29:14)
Yeah, I think that's our biggest fear going into this is that we are developing all of these tools and what's going to happen is that they'll be put out there, but everybody just goes back into their old habits and does it the old way because that's what's comfortable when things get heavy and busy, that they'll just kind of go with what's comfortable. and I think it's going to be like any type of systems change that you have, you know, it really just involves, talking about it, getting it out there, showing them that it works, and then, you know, just making that part of almost part of your SOP.
Eventually, so yes, we do struggle with that. I mean, a lot a lot of this, you know, a couple of years ago, most people would just be like, I don't know what you're talking about. This is too much, this is AI, this is crazy. but now we're at a point where like some of those people who I had originally thought would kind of just steer away from this are really you know leaning into it because, like, my god, that saved me five hours. You know, I I would have spent all week just kind of going back to that. So I have faith that if you're
continuing to praise this and show them things and you know be in a place where you can just demonstrate that these tools are working, then you know we're dealing with right with this with our scheduling team right now where we have started using an agent to make calls and you know help to you know fill visits for us. But I think a lot of people are hesitant to kind of step into that because it takes a little bit of time to learn how to use the tool and then to change your whole process when you know that it's taken you have to make you know 500 calls
Calls a week to get the all of the visits filled, it's gonna take a lot of time. So I've honestly have stepped right into it and started using the agent and showing them, hey, this is this works. And you know, we're at a place where I've decided I'm just going to do this for maybe the next month and say, okay, I can get the entire weekend filled on Monday using the agent. See, it like if you if you take a little bit of time and kind of really immerse yourself in this, you can be at this point too. And I'll, you know, I I think a lot of it is just me, helping them along the
the
way and embracing it and also kind of just showing the possibility. So yeah, it it's it's it's back and forth, but
Miriam Allred (31:17)
And it's it it sounds like that's
working. You know, think of your most like tech resistant colleagues. When we say people like us, we say to them, it'll save you time and and effort and productivity. We
Andrew (31:29)
Mm-hmm.
Miriam Allred (31:30)
say all these things, but what you said a second ago, a lot of these people like they're comfortable and they've been doing this a long time and they're just like they're comfortable in like the way that they've been doing things. And so it's it's hard to rock the boat. And we've talked about this before, like.
We're weeding some people out that just aren't keeping up with the times and aren't gonna take our companies into the next five and ten years. And that's a hard
Andrew (31:51)
Yeah, yeah.
Miriam Allred (31:53)
balance to strike right now of like we don't wanna let them go because they have all this like experience but it's like if they can't adopt and embrace these things fast enough, then we have to like
Andrew (32:04)
Yeah, yeah.
Miriam Allred (32:05)
think twice. Do you feel like you've been in that balancing act
yet?
Andrew (32:10)
I think we're almost there and
John Sneath has done an amazing job of kind of letting our company know: hey, we're not in this to take your jobs away. We want this to be transformative. We want this to make the quality of your day-to-day much better. And I think, you know, I'm an example. I jumped into AI, I started using it and was able to kind of transform what I was doing day-to-day into something that I feel, great about and, you know, I want to grow. So we're really trying to kind of give that message to everybody in the company. Like, maybe not necessarily leaning into AI.
But leaning into something you feel passionate about that does something for the company. And you know, maybe scheduling an AI isn't your thing, but you're great at building relationships with caregivers. That human component is never going to go away. We need to be able to talk to our caregivers and understand their challenges and everything they're dealing with day to day. So I think it's really just about people. Okay, I can accept that AI isn't something I'm good at or that I'm interested in, but you know.
I'm gonna put myself in this position and really start to try to grow here. And that that's really the environment that we're trying to kind of, you know, grow right now at Tribute to deal with this.
Miriam Allred (33:21)
And something that
we're aligned on is for the foreseeable future, you need a human, human judgment
Andrew (33:28)
Mm-hmm.
Miriam Allred (33:28)
using all of this AI. Like I I would imagine, do you have anything that you've built that is like one hundred percent automated, like a human is not even in the loop, or is is everything that you've built still so far there's like a human pressing the green light at the end of the workflow?
Andrew (33:44)
I'm at a point right now where I could press a button and have the human completely taken away from it, but I am not confident in that yet. And I don't want to be in a place where I wake up in the morning and there was a hallucination that occurred and everything that I've built kind of just loses all of its trust. So no, I'm very f fragile with that and I wanna make sure that we have tested, tried and true, and but I mean, but at the same time, I mean you're never really gonna be at a place where you can fully schedule without a human. I
All of the kind of relationship building we do here at Tribute, and I'm sure many other companies.
If the agent gets on and they say, Can you take this visit from, you know, seven to seven in Baltimore with blah blah blah, they're gonna I think caregivers are gonna hear that and be no, I'm busy, I don't wanna deal with that. But if a human calls, you know, one of our schedulers, Jen or something, you know, she says, you know, hey, how's it going? How's your family? Like that right there is so much more value to a caregiver. And I think they're gonna, you know, want to say yes so much more than an agent. you know, and we're not they're never gonna going to be able to build relationships with any of these people. So
So yeah.
Miriam Allred (34:52)
Did you say though you are
starting to deploy voice in certain areas? Are you? Where
Andrew (34:55)
Yes, yes.
Miriam Allred (34:57)
where are you starting with voice?
Andrew (34:58)
so we started in Massachusetts. actually Viv created a voice agent for their they call it their OVM platform that they have within scheduling and we have found that
If we start on a Monday to try to fill the visits for the weekend, that we are seeing the most success there. And it's really just something that we never had the space to do. I mean, we've always,
Come back on Monday, everything's kind of, you know, not necessarily a mess, but everything's kind of just, you know, changed from Friday to Monday on the weekend. And so you're kind of left, kind of having to clean all that up. and so you know, this idea of scheduling for the weekend on a Monday is like, my God, this is great. So we deployed all of the voice agents for that and saw a huge success. I think we filled 75% for of all of the open visits just using a the AI agent for.
that and we got the entire weekend cleared in Massachusetts. I think it was the first week that we used it. so that was a huge, tremendous value to us. Yes.
Miriam Allred (36:02)
So it's calling caregivers. It's
talking to the caregivers. And sorry, to make sure I understand, you're saying it's Monday to fill the following weekend. So
Andrew (36:10)
Mm-hmm. Yeah.
Miriam Allred (36:11)
the week is basically covered. Like the week is pretty stable, but you're saying on Monday we already predict like availability and call out like issues arising for the following weekend and we want to get ahead of it on Monday. Whereas yeah.
Andrew (36:22)
Well I think yeah.
so I think
Part of the issue is, you know, because we have such a high utilization, we're often left with holes for Sunday. We just don't have enough Sunday caregivers to fill an ongoing care team right now in a specific place. So week to week, we're in this place where we're just having to fill the weekend. it's always been a thing here, and I'm not sure about other companies, but you know, usually what's happening is that you know, still during the week we're dealing with call outs, dealing with new clients that are coming in, and then we're left, you know, waiting until
Thursday or Friday to fill those weekend visits, and those caregivers had already made plans up until that point. So for us, this just brand new territory to be able to start thinking about the weekend on a Monday. And so we have found by using the agent to do that, it filled the majority of those visits because the caregivers realized okay, now I can actually plan for the weekend, and it's been a huge value to us. So yeah.
Miriam Allred (37:18)
Okay, so Viv built built it, not
you guys, but I'd imagine you did a lot of like the the research and the prep and the refining with them. Did you have like a pretty tight
Andrew (37:27)
Yeah, so
Miriam Allred (37:28)
hand on it?
Andrew (37:30)
yes, Viv kind of you know, we started out as I think one of their first clients.
And they meet with us weekly and when they started rolling out their AI agent, it was something that we were both interested in. So they kind of built with us and got us to a point where you know, we've got something now that has really been almost customized for us here at Tribute, which has been great. so yes, they did build it. We are on the side building our own kind of like independent agent just to see if there's any more value there and to, you know, start thinking a little bit more about that automation.
But yes, they they did build it for us.
Miriam Allred (38:08)
And what QA did you put in place to know that it was okay? Like obviously you're nervous, like to literally like turn that thing on. I just I still think it's so scary to turn that thing on and then all of a sudden it's just like
Andrew (38:17)
Yeah.
Miriam Allred (38:18)
calling your caregivers. Like that is a huge step. But what QA did you put in place or did Viv put in place to ensure quality assurance on that?
Andrew (38:29)
So
lot of a lot of testing. I think they released it last year and we really didn't start using it until May. So you know I'd go on, I'd put my phone up and have the agent call me, see if it was saying the right things and just a lot of trial and error of going back and forth and making sure that you know when that agent did call the carrier it was giving them the right information. Because you know we don't want somebody saying or the agent saying, you know, we're gonna give you $500 to take this visit or something like that where it just messes everything up. So yeah about a year of quality improvement where
working
with Viv, going back and forth. and then we trialed it in May in Massachusetts and we are now trialing it for the entire company still. we're not a hundred percent we haven't a hundred percent rolled it out, but we feel pretty confident that we'll be there pretty soon.
Miriam Allred (39:16)
And I I guess I see it as twofold. Like it's one thing for the AI, the agent, to have enough information and to have all of that information be able to like speak to it. To me, that's like table stakes and we're there already.
Andrew (39:26)
Mm-hmm.
Miriam Allred (39:28)
But it's the tone and the personality and the empathy and the the conversation that still feels a little clunky, a little like AI.
Andrew (39:38)
Mm-hmm.
Miriam Allred (39:38)
Like it's getting very good. I have heard extremely good agents,
Andrew (39:40)
Yes.
Miriam Allred (39:41)
but you and I still can like tell the difference.
Andrew (39:44)
A hundred percent,
yeah.
Miriam Allred (39:46)
Does that worry you? Does that bother you? Does that br bother caregivers? Like what do you think about that?
Andrew (39:51)
We were pretty cautious about this. You know, we had sent out a ton of you know
It's just emails and you know, we s we created some videos sending out to our caregivers that hey, we're doing this not only to, you know, support our scheduling team, but also for you. Like, you know, we want you to be able to plan your week accordingly and to be given offers when we don't have the time to be able to do that. So, you know, we try to be really delicate around that process and you know, thinking about like, hey, we don't want these agents calling people at eleven o'clock at night and all that. but yeah, you're right. I mean, the agents themselves are not humans and
I think when they're calling, they're not reaching them to that level of scheduling partner. And I I don't think it will ever be at that point, to be honest with you. I mean, unless we really get into this call intelligence platform that we're building and it knows exactly how to talk to this specific caregiver. Like that, it's just I can't I can't imagine AI ever being there. So
Miriam Allred (40:48)
So we're
all just gonna get used to it and just accept that we're talking to AI. Like, do you cause my qu my quality bar for this is like
Andrew (40:53)
Yeah, yeah. Mm-hmm.
Miriam Allred (40:57)
if I would be comfortable having a conversation with it. Most AIs that I talk to with the bank and restaurants and all this stuff, it's still frustrating. To me, it's still frustrating.
Andrew (41:05)
so frustrating.
Miriam Allred (41:07)
And so if I were deploying it in my own business, personally I wouldn't be yet because but I don't know, you were just saying like if you tell the caregivers this is AI.
You you kinda like preface all of this, but explain the advantages of like speed and scheduling and I guess if you explain all those advantages, like then maybe we just accept that we're talking to AI and it's not that great, but the advantages outweigh just the lack of tech, you know, quality.
Andrew (41:35)
Yeah, yeah. And
I think we're somewhat there. It's it's yes, there are advantages there, also disadvantages because you're not going to get that quality. Yeah, we all all hate talking to a bank now to be able to get those stupid answers that they never you know they never give us what we want to hear. So you know, we're still at a point where even if they do say yes, we are calling them up and it's saying we're going to connect you with, you know, so and so from our scheduling team, to just talk about details. So no, I'm not a hundred percent convinced that we are
anywhere close to having a fully automated scheduling team anytime soon. I think it's just going to be around that support with making those calls. And then once we do find somebody, we can connect with them, have them speak to a human, go through the final details. you know I I do think that there are platforms out there that right now that are doing it and I'm hearing some success. But for Tribute when we place so much emphasis on our caregivers and making sure that they have all the support that they need the
you know, we're we're a hundred percent going to need a scheduling team for the foreseeable future. So
Miriam Allred (42:38)
But it sounds like
you have a pretty quick live transfer then. Like it's really just getting them on the hook, like answering their questions, like kind of just the the quantitative information. And then if they say yes,
Andrew (42:48)
Mm-hmm.
Miriam Allred (42:49)
then it's very quickly a live transfer for them to kind of finish out all the details with a human.
Andrew (42:51)
Yes. We set it up we
set it up with Viv to make sure that they we had that ability. So they'll say, Yes, we're transferring you to a scheduling partner right now and then somebody from the scheduling team will pick up the phone and confirm the visit with them. Yeah.
Miriam Allred (43:04)
Okay. Are
has Viv or you guys built like a a parallel text based system where you're sending out shifts and having having AI have text based conversations with caregivers simultaneously?
Andrew (43:16)
so no, Viv has kept it all in-house within their platform. and I think we like it that way. we've been talking a little bit more about like, hey, yes, we do need kind of a space, you know, where we're not just going into each individual visit for that because we're gonna need that. We can't just, you know, keep our eye on every single AI conversation in a different window everywhere. So we're getting there, not quite there yet, but you know, some of the stuff that I am testing right now shows that we can have it in one environment and that we're
works and so yeah yeah.
Miriam Allred (43:47)
Okay. Yeah. The reason I
ask is like, yeah, call volume versus text volume. I guess you assume and think like a lot of people are more receptive to text. Like granted people answer the phone too and you can have more of like a flushed out conversation on the phone. But like if I were an agency monitoring my call volume versus my text volume and deploying like automated texting first before the voice
Andrew (44:07)
Mm-hmm.
Miriam Allred (44:07)
stuff is the approach that I would take because to me that's like a little bit like safer foolproof, like trusting an AI to to carry out conversations.
Andrew (44:16)
Because you could see, I mean, yeah, I
Miriam Allred (44:16)
Via text versus phones.
Andrew (44:18)
mean, we went the opposite way. We s part of our policy has always been call caregiver first. Call, call, call, and if you can't reach them, get the get a text. You know, you send them a text. yes, we do get more responses via text, but I think the quality of the caregiver's experience and even our experience has always been just that call. So, you know, that's what we're doing now with the agent. And we see the transcript happening in live time. So if it makes a you know a mistake or whatever, you can just quickly end that call, call the
caregiver
up. but yeah that I I do see what you're saying and I think that you know texting allows us to have a little bit more control over the situation but we really want to stick to what we believe kind of is the right approach as well while we're going through this. So yeah.
Miriam Allred (45:00)
And sets Tribute apart. You know, every other agency is texting, but if you guys are calling
and having a conversation, like that stands out to the caregiver and differentiates you. I I wanna talk
Andrew (45:07)
Mm-hmm. Mm-hmm. Exactly.
Miriam Allred (45:10)
about Claude and Claude Code. You know, we've all seen it, heard it. I would say I'm actually personally like behind the game on some of this.
Andrew (45:17)
Mm-hmm.
Miriam Allred (45:19)
that came out, what, has it been six months even or not even? It's been what
Andrew (45:23)
They've had it for a while, but I think this whole vibe coding kind of phenomenon is really more within the past six months that really people a lot of people have latched onto it. So yeah.
Miriam Allred (45:31)
Okay. From a from
like a learning standpoint or just like the ability to use it, is it a huge jump from like what people are used to and Chat GPT and Claude and other, you know, prompting software? Like is this a huge like learning curve to be able to navigate that and execute in that or no?
Andrew (45:52)
So I yeah,
Miriam Allred (45:53)
Set a load of questions.
Andrew (45:55)
no, I everybody in my company is very interested in claude code because they see me on it all day long and they see how fast I'm able to generate some of the requests that they have. And so I was like, yeah, it's pretty easy. You just you type in it just like you do with Claude on, you know, the regular Claude platform in your app or on the internet. and so I
demoed how to use cloud code in front of you know a big portion of our leadership committee and everybody was like, whoa, this is too much. They saw terminal and they were like, I don't even know what terminal is. I don't know how to do that. and I
Miriam Allred (46:28)
Define terminal for everyone listening.
Andrew (46:31)
so terminal is kind of I don't know if some of some of the old timers out there remember like DOS or DOS shell on a Windows computer, but it's kind of the you know the backup kind of like this is where you enter prompts for your computer type thing.
It's on every Mac. You go to up to the top screen and you just type in terminal and it will come up as a black screen where you can kind of just put in stuff. So it's very unfamiliar to a lot of people because we don't use nobody uses terminal day to day. And I think even you know, I don't use it, but I think a lot of people use Claude Code within the Claude app itself now, too. So I have not done that myself, but I think it's really the same thing, but you're just not having to go into terminal. so as far as being able
able to use it. It's very similar to like a cowork and
You know, if people aren't familiar with cowork is it, you know, you usually have two kind of functions within Claude where you can just have your chat and then you can have cowork, and cowork will work within a specific folder on your computer. and then with claude code, it's really just full access to your entire computer. You can prompt it to go on to your browser and look at specific things. So for me, it was just kind of like a natural evolution. I was like, I'm doing all this work on Claude.
chat all the time, but I need to have multiple windows open and things going on at the same time. And this allowed me to do that. So I could, you know, get my job done as, you know, scheduling director for a while and then in the background have, you know, five different things that I'm building and just have that be in terminal. Yeah.
Miriam Allred (48:05)
So does it mean so again, I'm I'm kind of like a beginner on this. Does it do you
mean that it's it is l referencing everything that's sh displayed on your screen? Or when you say it has access to everything, you literally give it access to all of your files and all of the data and information on your computer?
Andrew (48:21)
Yeah, I mean you can be very specific about which files you want on there or, you know, what access you want it to be limited to. But essentially yes, you can have it just access everything on your computer
But as far as prompting, you know, you talk to it just like you would, you know, your regular Claude that you use day-to-day or ChatGPT or anything like that. So literally, you're like once you have it open and Claude is up and running, you can be like, I want to build a dashboard, and then it will walk you through how well what dashboard do you want to build? And then it does all of that. I think where it becomes very complicated is once you start getting into like APIs and building an MCP server and all that stuff that we're not thinking about day to day. But I think
What I've been trying to tell everybody, and we have a few users or a few office associates that are very interested in wanting to learn it, is that just take slow steps, you know, ask it to build a dashboard, see what happens, ask it, how do I connect to Viv, see what happens, how do I connect to DocuSign or, you know, any of the other companies that you're using day to day. Most platforms have a direct API connection now that you can use. it's just a matter of how much time and how much, you know.
you really want you wanna do this and learn about it that it's going to kind of be the differentiator. So yeah. Yes. Yeah.
Miriam Allred (49:36)
Okay, but that was a good filter where you started, which was you showed it to the whole leadership team and they're like, Well, this is awesome, but we're not ready for this. Like
it's easy to use, but what what's the concern there? I guess what was like John's concern? Like people can get carried away or we could build things and not not have the QA, therefore it's just getting like
Andrew (49:56)
Yeah.
Miriam Allred (49:57)
murky and messy. Like, is that the concern?
Andrew (49:59)
So I think for me personally, and you know, this is something that I've talked to John and others, that it's one thing for one person to be in there and building something, it's another if everybody's on your API connection and you know, trying to change things, you're dealing with private information, things like that. So I think if you are going to be considering doing this for your company, it's really important that you z you designated, you know, a few individuals that really understand what they're doing day to day and connect with each other to kind of make sure that you're not messing anything up.
Miriam Allred (50:29)
What's
your advice to everyone listening to this? Do you feel like if people aren't just tinkering in Claude Code, like they're about to really get left behind? Is it that extreme?
Andrew (50:42)
No, no, you could still tinker in Claude and build from there. I mean, a lot of the initial work that I had done was within the Claude chat.
It's just, you know, it it it really just becomes which platform do you feel most comfortable. You can get a lot more done in terminal, but still at the same time, you can build and it's just a more manual process within Claude Chat. So you're not gonna be behind. It's just you know, it really just depends on s what you want to do and achieve. you're not going to be able to kind of adjust, settings within all of your different platforms within Claude Chat, but you can still build and you can, you know, what I ask a lot of people to do is when they have an idea.
About something that they want. I ask them to go to Claude and just say, and just talk about what you want to build, build that dashboard, and then send me a synopsis of that chat that you had. And then what I'll do is I'll put it into terminal and say, here's what they're thinking. Let's just run with it. Let's see what happens. And Claude Code just kind of helps me build from there. So yeah.
Miriam Allred (51:37)
That's great. That that's that's
a perfect example of like people can use it kind of like in a light version. They know what they want. And so they
Andrew (51:45)
Mm-hmm.
Miriam Allred (51:45)
can have the conversation and ask the questions. Like I love what you said a second ago of like asking it, again, this is like getting into the advanced territory, but asking it how to build an API or how to connect an API with another tool that you use. Because you and I, you know, I don't know, five
Andrew (51:57)
Yes.
Miriam Allred (52:00)
years ago it was like, I don't know how to connect an API, but now it's like you ask Claude.
Andrew (52:02)
yeah, it's nonsense.
Miriam Allred (52:05)
How to connect X software and Y software, and it will tell you how. Granted, it's still a little bit complicated. You have to like go and find and hunt down these API docs and information
Andrew (52:13)
Mm-hmm.
Miriam Allred (52:14)
stuff, but you just ask it what you want, tell it what you want, have that conversation. But I like this concept of then you pass it to someone like yourself that's like
Andrew (52:22)
Yeah.
Miriam Allred (52:22)
qualified to like take it to the finish line or over the finish line, whereas, you know, they don't have the time or the skills probably to do that.
Andrew (52:31)
And I keep telling everybody too that anything is possible right now, like when it comes to what you want to build and wanting to make your day easier. and if you kind of just take away like, that's too hard or that's too difficult, if you take that piece away and really just start just putting out all of your problems, just start with the problem, put it in there, see what can come out of it, and you can really, really get to a place where you've built something that you can use day-to-day. So
Miriam Allred (52:58)
Do you worry about people wasting time on this? Is that a concern? Like people are busy and they have priorities. do you give them free reign and do you worry about people wasting time on this?
Andrew (53:08)
Not necessarily. I mean, I think if you're not focused on something that's going to make your life easier or you're just using it to write an email, yeah, I mean I think that's a waste of time because you're just you know, having it do the thinking for you. It's really just using it as a tool to kind of take a thought you've already had and go forward with it. So we haven't run into a problem where people are wasting their time on it that I'm aware of yet.
And, you know, I think if people are really interested in wanting to get this done, they're gonna go after hours and really start, you know, going onto YouTube and, researching how to get this done because once they see the full potential, it's really gonna open up their eyes a little bit more to it.
Miriam Allred (53:50)
It seems like most of the success people are having is all self taught. It's all self learning. It's literally just like spending time with the tool. Have you yet I guess have you spent a lot of time on YouTube? Are there other resources or educators or people or places that you have gone and found valuable information to give you a leg up? Or is it more just the self learning?
Andrew (54:11)
You know, I think it's a little embarrassing, but you know, I
I tend to Doom Scroll in my screen in my free time where I'm just on Instagram going through those videos. And if you start kind of like liking the AI ones a little bit more, that's that's been helpful for me because it's
Miriam Allred (54:24)
It's gonna give you more
Andrew (54:26)
kind of like giving me both at the same time. But y I think what happens is like a lot of the times when the new update comes out, because you know, that can be pretty valuable in your day to day, I'll go on and be like, this new fable five point one, what's going on with it? or you know, for me right now too, my big thing is cybersecurity because you know, we are using a lot of personal information.
and making sure that we're making the best choices around that. so you know, things that are important to my function right now, yes, I think that's been important for me. But as far as wanting to expand and grow upon AI, yeah, I mean Claude has a ton of free courses that you can just jump into. I know somebody from our company that started using that last week and they're getting a lot of valuable information there. So yeah, the stuff is out there and you know it would do no harm to kind of just, you know, start using that to learn more.
So yeah.
Miriam Allred (55:15)
Dive in. Yeah. And everyone, I
I guess it's just like figure out where you are. And everyone I don't know, to me, even me, like this is self-serving, but it's just like, you know, taking a step in like the right direction, like exposing
Andrew (55:26)
Mm-hmm.
Miriam Allred (55:26)
myself to something, trying something. I I don't know if this is just like so generic to say, but like the the world is changing, you know, like a lot of the podcasts, the stuff that I consume, it's like we're really gonna start to see and feel the impact 2027, 2028, 2029. Like
This is just
Andrew (55:43)
yeah, yeah.
Miriam Allred (55:44)
the beginning and it's about to like really ramp up. And that's scary. But I hope everyone should feel some heat on their on their job, on their company, on their growth. Like everyone should feel this like healthy, good pressure.
Andrew (55:58)
Yes, yes.
Miriam Allred (55:59)
And it should like we talked about, you know, like turning nerves into excitement. Like we should feel some excitement of like this is pushing us to be better. And we all need to like get on this AI train whether we want to or not. Because
It's gonna impact it's already impacting our day to day, but it's only it's only gonna continue.
Andrew (56:17)
a hundred percent. I mean I I'm always thinking about like late eighties through the nineties. Like think about an accountant, like and how different that job became and what it is now with a computer today. Like that really it's just another tool. We're all growing with it. We have to learn how to adapt and if we can use it to make our lives better, then that's really how we have to embrace it and we have to think about, you know, how can we increase the quality of our day-to-day using this tool. So
Miriam Allred (56:46)
Yeah.
Andrew (56:46)
Yeah. Yeah.
Miriam Allred (56:47)
just a couple c questions in closing here.
has the biggest unlock for Tribute been embedding AI between your systems? Like there's
Andrew (56:56)
Mm-hmm. Yes.
Miriam Allred (56:57)
so much data and documentation and information transferring in home care. Would you say that that has been like single handedly the best use?
Andrew (57:08)
yes.
I in sneak peek to HCAOA. you know, we've developed this one tool around lead taking where, we were, you know, the lead would come in by phone, you'd have to take your notes, you'd have to go into Teams to write to the scheduling team, you'd have to get the contract and docu sign, then you'd have to go into HubSpot. We found that if we just put it all together into one tool and have it spit out all the information at once, it saves you an hour of time that you you know, that you would be doing every day. So yes, a hundred percent that is where I think the
full potential is in home care right now. and I think if anybody's thinking about doing this, that that should be where your focus should be at. So yeah.
Miriam Allred (57:46)
Yeah, I think that's where people
again, it's the advanced skill set, but that is the biggest unlock right now is how are we connecting payroll with scheduling with CRM with DocuSign? Like you have your kind of like
Andrew (57:58)
Yes.
Miriam Allred (57:59)
core tech stack of five to eight platforms, but
Andrew (58:02)
Mm-hmm.
Miriam Allred (58:03)
unfortunately none of them are talking to each other and they all need to be. And now with AI, we have the ability to get them all to
Andrew (58:10)
Yes.
Miriam Allred (58:10)
talk to each other and to transfer that data and information.
Andrew (58:13)
A hundred percent and like it's we all do things differently. And that's why I think it's very important if you have somebody in-house, maybe that's not even fully fluent on how to use AI, to just lay out the problems to somebody that understands operations, that knows that this is what takes the amount of time each day to get this done, and then tr either do it yourself or translate it over to somebody else. So
Miriam Allred (58:35)
And you guys have done it in house because you can do it faster than anybody else.
Andrew (58:41)
We're not having to wait each week to get to a meeting to talk about our problems. We can just, you know, do it each day and kind of work through it. So that's
Miriam Allred (58:48)
Yeah.
Andrew (58:48)
yeah, that's such a tremendous value to us.
Miriam Allred (58:50)
And and to end
on just like the where why this all matters is like talk about the team satisfaction. Like talk about your your whole administrative team. Like has job satisfaction gone up because of these tools and productivity?
Andrew (59:04)
I think we're starting to see it right now. you know, our finance team c in July they're gonna be running it in a few days.
ran their entire financial statement by the press of a button and something that took them you know two days to to do like it that just the the stress of you know that work and having to get that done each month is completely gone. So yes I think people are starting to feel relieved and satisfied knowing that you know some of this more cumbersome, you know repetitive work is now being done for them and they can focus on, more of the valuable things. So yes.
Miriam Allred (59:37)
And that's gonna be interesting to see. You know, people might ask, like, what do I do with all this time on my
hands? That's we've never had that luxury in home care, ever. No nobody has ever said, I have more time on
Andrew (59:45)
Never, never.
Miriam Allred (59:48)
my hands. What do I do at that time? We are crossing into that territory of like, what is the best use of your time? Because now AI can do a lot of the the manual tasks that you were
doing beforehand.
Andrew (59:58)
Yes.
I still think we're going to be in that place forever because somebody's always gonna find another hole somewhere else and you know we'll we'll figure it out. But
Miriam Allred (1:00:05)
But I think back to
calling the caregivers. I think of like taking off a lot of the manual, minutiae to be able to
Andrew (1:00:11)
god, yes. Yes.
Miriam Allred (1:00:13)
talk to more people. Your finance, you know, this is far-fetched, but like your finance person, like having a conversation with a client or a family member, like just being more proactive and driving more conversations because you have the the time to do so.
Andrew (1:00:25)
Yes, yes. I can a hundred percent agree there.
Miriam Allred (1:00:28)
What about like growth, hours, revenue, just like the trajectory of Tribute? Have you guys already seen
Growth strides because of this or do you feel like that's coming?
Andrew (1:00:39)
you know, there are a few factors here, but we've seen an impact on our gross profit this year already. you know, we've broken it down to the amount of time that we're saving. last time I charged we had a presentation on it last week, so around a little over three hundred hours each month that we've saved from some of our AI tools. And that's, you know, a couple full time roles that we otherwise would have had to hire. So we are seeing the impacts and I think it's only gonna get greater going to get greater from here. So yeah.
Miriam Allred (1:01:05)
Yeah, definitely the internal, like time and
cost savings and it'll be interesting to see as that then translates externally. Like, you know, we're cutting costs and saving time, but then also growing at, you know, X margin. Like I think that's
Andrew (1:01:17)
Yeah, growing but also increasing the quality, you know, and the pay of our some of our caregivers and things like that that really matter, that allow us to, set us apart from, some others out there. So yeah. I love these type of comics.
Miriam Allred (1:01:27)
Yeah. Andrew, this has been awesome. I know this has been kind of just like a raw, unfiltered conversation. I've strung you along. But this is
just like an honest real take of like somebody building it in the trenches, thinking about this. Like I just I started at the beginning of like, I'm tired of thinking about this. But then when I start talking about it, I'm like, there's so much more. And everybody needs to be thinking about this and talking about this and auditing their team and their processes and
Andrew (1:01:49)
Mm-hmm, mm-hmm.
Miriam Allred (1:01:51)
tinkering internally and finding someone internally that's kind of like passionate and has like a knack for this.
I think every single home care company needs to just be find that person and do the tinkering and do the research and do the auditing or like you said, lay out all the problems, lay out all of the manual processes and then even take that to someone and like exp
Andrew (1:02:11)
Yes.
Miriam Allred (1:02:12)
express that and explain that. Like that's the stage that we're in and that needs to be happening. And that's why I wanted to have this conversation.
Andrew (1:02:16)
And I can guarantee yes.
I can guarantee there's there will be somebody there for you that you know wants to do this. So yes, yes.
Miriam Allred (1:02:23)
So find them, seek them out and make it happen. And then for all of
us, just like, you know, listening, it's like find where you are and then take the next step to get to that next level. Like we all just need to keep leveling up. And it doesn't matter where we're starting from, but just like inching forward.
Andrew (1:02:36)
Mm-hmm, mm-hmm. Exactly.
Miriam Allred (1:02:37)
And I know I'm re- relaying all this advice. Any other advice for everyone listening that you want to impart or just anything you've learned that you want to share?
Andrew (1:02:43)
Yeah just
just throw yourself out there and look at your problems and there there's going to be something that you can solve. So that's that's really all I can offer right now. Thank you for having me, Miriam. Same.
Miriam Allred (1:02:50)
Awesome, Andrew. Fantastic. Well, thank you for joining me for this conversation in the lab. Andrew, I'm really excited for HCAOA in a couple of weeks for anyone listening
to this, you know, recording first week in September, HCAOA's end of September. But we're gonna put a visual to all of this. So Andrew's actually gonna show a lot of the things that they built and we're gonna get into Claude and all of that good stuff. So this was kind of a precursor to that. So I'm looking forward to it. So thank you for being here and thanks for listening, everybody.
Andrew (1:03:14)
Thank you, Miriam.