The Companies Winning Today All Have This Advantage With Founder & Co-CEO of Avoca

But I think where AI can actually have the most direct value add is actually in the, in the blue collar, in the physical services. And so that's really what we're excited about. And it's not about replacing people. I think that's a very common misconception narrative. Obviously, there will be some people that will have different jobs. Instead of taking 10 years to go from $1 to $100 million in revenue, we want them to be able to go $1 to $100 million in 2 or 3 years.

Ladies and gentlemen, welcome back to the Unemployable Podcast. I'm Jeff Dudan. If you are a 2017 MIT grad who built a billion-dollar business revolutionizing voice AI and putting it to work in customer service centers everywhere, your name can only be Tyson Chen. Welcome, Tyson.

Thank you so much for having me, Jeff. I'm really excited to be here.

Yeah, excited. We've been talking about this for quite a while, but you know, when you're, when you're building a state-of-the-art, fast-growing, sought-after, in-demand company called Avoca, then like it's hard to get you on the calendar.

Oh, you're, you're a very busy man yourself, but, but appreciate it, Jeff. So, so glad to be here and to also be working with you guys.

Yeah, 100%. Here's the opening question. I've heard you talk about having a human in the loop. What do you mean by that?

Yes. So human in the loop is actually pretty simple. Basically, the idea being when the AI struggles. So let's say our one of our AI agents is handling a phone call for, for a customer and there's something that requires a human touch. In an instant, one of our human operators can instantly take over the call and be able to, to bring that to fruition. And that's actually a concept that I first discovered almost 5 years ago when I was actually working in Silicon Valley at a company called Nuro. Nuro was building autonomous vehicles for goods delivery, and I was actually on the team that pioneered the human-in-the-loop system for Nuro in the sense that when an autonomous vehicle was stuck on the road, let's say there's, you know, one lane is closed, and so you have to— all traffic has to go through one other lane. Well, that's an experience that there's just not much data on. And so a human, uh, uh, what we call a guardian, would beam in and take over the vehicle. And so, um, kind of took a page out of, out of that playbook, uh, for Aboqa as well.

Okay, so Did you get that concept? Uh, did, did, is it, what's the name of the company? Noru?

Uh, Neuro. Neuro. Neuro.

Did they get that comp, uh, from Guardians of the Galaxy where they had all of those people that were remote controlling the spaceships from a, a huge, uh, center? You know what I'm talking about?

Have you seen that? Yeah.

Yeah.

Uh, I have, uh, it might, it may have been that one or, or maybe something, something else, but, uh, but yeah, it was, it was always. It's funny, it's like this, we always joked about it. It's kind of like playing a game of like a racing game where the stakes are up all the way because when you're controlling the car, like you don't get any of the actual feedback. It's actually one of the hardest jobs I've tried it, like not in actual real life, but just testing it. It's like you're literally playing a video game, but you cannot do a single mistake. And if you mess up, you know, there could be serious casualties on the line.

Right. So, um, so there, there would be a, a room full of people that are in simulators and they would be, be waiting for an opportunity to come live and then it comes into them and all of a sudden they're, they're in a live action scene. Uh, scene.

Yeah.

And, but yet they don't have the ability to shoulder check.

They, they, you know, they probably don't have, uh, they have cameras, like they have everything on the screens and like a Guardian, they call monitors, but exactly. They don't have the sensory and like real, real life feedback. So it is very difficult and, and they're extremely highly trained. When it gets passed to them, they do see the context. And it's like, it's like 5, 4, 3, 2, 1, and then you're in. And then beforehand, they get to see like everything that's happened.

When they get the, when they get into the ride, is the car always stationary?

Yes, that is a rule. So they're not, otherwise it's just too difficult.

Yeah, yeah, you're going down the highway and, oh, here you go.

Yeah, exactly.

Yeah.

So you're So you're at that company, um, you've, uh, you're, you're 9 years outta school, uh, now. Um, what was a skill or specific experience that you had at that company that turned out to be the most relevant to creating Avoca?

Hmm. Yeah. Uh, love that question. I think that I'll give you, I'll give you an experience from Neuro and then maybe, uh, an experience from, uh, BCG as well, which is my first ever job. So I think from Neuro, the most valuable skill, because I worked on a team that was helping bring like products to the operations teams. And I think one of the biggest things that we learned, that I learned is like just how important it is to actually, you know, get the buy-in and like learn how to build great products from the people that are in the fields. Um, I think that's something that has carried over in a big way to Avoca where, you know, we work. I think part of the reason why Avoca has built such an amazing brand and, and, and are, and are the market leaders in this space is because, um, we deeply understand um, you know, our customers, and we get in the weeds with them. And so, for example, we, you know, at Nuro, there's a, uh, an app I built, uh, for helping, uh, autonomous vehicle operators to map. You know, in order for autonomous vehicle to be able to operate in any, um, uh, geography, you have to do this like mapping and localization exercise where, uh, someone has to physically go and just like like they basically drive every nook and cranny of every street in a, in a particular radius. And we built an app that essentially helped them to do that far more efficiently is one of the things that we built on the team. And yeah, I remember like for like 2 or 3 straight weeks getting actually in those vehicles, learning every single part of that operator's day to day. So that we can, and then team up with them because a lot of times, you know, these guys are actually really smart. They've also figured out little hacks and tricks and stuff. And so like really, you know, having that type of in-depth product development, I think we've taken that same concept and that's a lot of the way we build new products at Avoca is we actually send in a team, you know, I go still oftentimes and and my co-founder as well, and we actually sit in, uh, stay like a week, two weeks with our customers, uh, and actually build alongside them.

Walk us through the earliest version of the product, or even maybe even before that, when did you first realize that there was an opportunity to win in this space? And then what action did you take?

Yeah, um, yeah, I love that question. I think the earliest kind of formation, um, really predates, uh, even, even, uh, so Bhav and I, we met, uh, back at MIT, but really the, the story predates that. Um, it actually started— I grew up, um, answering the phones for my mom's acupuncture business, which is one of the first ever acupuncture businesses in, uh, Pennsylvania. And I always remember my mom would say, you know, wherever you are in the house, Tyson, as soon as you hear the phone ringing, stop what you're doing and chase that phone down because the sound of the phone ringing is the sound of money. Apoorva has also a very kind of a similar story. He answered the phone for his mom's business in Michigan. And so I think even back then we were, you know, we had the remnants of it. We had like, you know, we'd gone through the job as kind of a one-man CSR army., and we realized, wait, this is very repetitive. This is something that if AI got good enough, it could, it could perhaps even perform this job better than we can. So that was kind of the initial ruminations. And then fast forward, I guess, up until around 3 years ago, so 2023, what actually brought us into the industry of home services was we were in Dallas for a restaurant convention because at the time, Avoka, our primary customer were restaurants. Like we had a few chains and we would answer their phone calls with AI. We'd take orders, we'd do stuff like, you know, make reservations. And so we were planning to get a bunch more restaurant customers. All of a sudden, one day, one of the restaurant owners introduces me to a buddy of his who owns a pretty big HVAC operation in the Dallas area, Rescue Air. And so I go in and instantly, you know, Josh and Michael are, you know, they're very friendly. They greet me and I show them the early version of the product. They're pretty amazed. They're like, wow, this is pretty cool. Like, can you change the voices? Yeah, I was like, I can change the voices to anything you want. Um, and then they were, they were just, you know, kind of blown away. And, uh, I think in particular, I learned about their operation and just how revenue-sensitive phone calls were for these HVAC operators. In fact, a single phone call, uh, could, could lead to a $30,000 or $40,000 install job. And then, then my mind was kind of racing at the time. I was like, what, wait, like For restaurants, you know, it's valuable. It helps save them some time, but a single order, if you miss it, is only like $30, $40. This is orders of magnitude larger. And then they told me about how much their current solution was, like, pretty shitty. Like, obviously the in-house CSR is pretty good, but after hours and overflow, which oftentimes are some of— are when the biggest ticket items come in, in times of emergency, they had just switched to their third answering service in the same year., and it's like, clearly something's not working. They're trying to fix this problem. It was like every single, uh, kind of checkbox was ticking in, in what like YC preaches as, uh, you know, things to, to look for in a really massive problem. Um, and so we kind of took a leap of faith, but literally, you know, a few weeks later, Apoorv and I, we deliberated and we're like, this is, this is worth going all in on. Uh, and so, uh, we, we put the, the restaurant folks, we still supported them, but we essentially said, the direction of the company, the vision of the company is moving towards home services.

And at that point in time, what were some of the hardest technical problems to solve with the product? And I guess if, you know, if I was going to double-click on this answer, it's, you know, would this have even been possible without AI? Or was the emergence of AI, because we've all, I mean, I've been calling call centers for 30 years, right? And they have IVRs, you know, press this button for this, press press this button for that, you have a recorded message on there, then it gets you somewhere else. So people have been trying to find efficiencies inside of call centers for decades. But your product, would it even have been possible to approach it this way without the emergence of AI?

Yeah, absolutely. Short answer is no. I mean, funny enough, during my consulting days, I actually worked on some chatbot implementations for Fortune 500 companies, some of the largest banks in the world. As I'm sure every listener would agree, those chatbots were utterly terrible and a terrible customer experience. No one wanted to talk to them. It was just like, how do I get to a human as fast as I can for many, many years? And finally, I think we're definitely early on this trend, but even in the GPT-3 days, we realized with the new, you know, large language models, something was fundamentally different. And, you know, and this could be, you know, have an amazing implication on all the voice applications. And so I would say the early days, you know, we kind of had that insight, and that's really why we started really pursuing voice agents. And there was a lot of challenges. I think in the GPT-3 days, this is even pre-ChatGPT, one in every five hallucination calls would have a hallucination. Like instead of saying, it'd be like, hey Jeff, you know, hope you're doing well. And later on it would forget your name and be like, and John, is there anything else you need? And customers are just like, what the fuck is going on? And so that's why, and that's why early days it was so necessary for us to have the human in the loop because the AI was so imperfect. There were so many hallucinations that we needed a kind of a failsafe to kind of go in and kind of save the day. And, and that's why we really— we invested like almost like 9 months into building the infrastructure that would support human-in-the-loop. There's a lot that's of engineering that goes into preserving the context of the conversation and having the tools so that when the human jumps in, they know exactly what's been said, what hasn't. And then I'll finish it off by saying It wasn't until GPT-3.5 that a lot of it got fixed. GPT-3.5 not only hallucinated significantly less, but also the latency was far better. So a lot of you listen to a lot of our early days, you know, it was a step function better than, you know, the chatbots that I had implemented, you know, in the, you know, 2017 to 2019. But it was still like very obvious, so very obvious that you were talking to a bot, that there was randomly long delays. It didn't sound super natural. And so 3.5 was a, was a big step change. And then beyond there, we just, yeah, we've also been making a ton of improvements. Now it's actually pretty remarkable. Uh, you know, it's, I think it's less than 5% of conversations, uh, under 2 minutes where people even realize that they're talking to an AI. It's, it's that realistic now.

You process an enormous number of telephone calls every day, interactions, uh, web forms, all, all of the, all of the channels of communication. So how important was it to be early in this space, knowing that now you've got this huge, continuously self-improving model based on all of the data that you're gathering and all of the experiences that your AI is having every day?

Massive. I think that's one of our biggest moats. I mean, now, uh, you know, in the last 2 years, there's been a lot more competitors that have rushed into this space and, you know, trying to, uh, because to start with, like, to, to build a simple voice agent, it's not too difficult. It's really that last mile and that last 5% of really optimizing everything that it becomes a pretty huge, like, engineering challenge. But to your question, I mean, I think that's probably one of our biggest moats, that we have been able to work with some of the top companies in the space early on, succeed with them, utilize that to improve our models and also the overall product., and, uh, and also just win the trust. Like, what we're doing, like, I think one thing, one kind of misconception is that a lot of our core stuff is, is customer service. It's actually not customer service. I mean, some of it is customer service, but really a lot of the, um, the calls and even the, some of the outbound stuff we're doing, it's very revenue sensitive. And so I almost, I always say that it's more, it's more similar to sales. Than customer service. And in sales, every single percentage point matters. Um, and so I think, uh, yeah, I think, I think that's, that's kind of a pretty big difference.

Maybe 10 years ago, uh, people of my generation, as we're building companies, we became very aware that we were hiring digital natives, people that, you know, I didn't grow up with technology. I mean, I had, I had a corded phone. I had 3 channels on the television. You know, we, we didn't have technology. So then when we started, when the workforce, the people that grew up on Facebook, they grew up with phones in their pockets and we started hiring that workforce. It was a, it was a shift in the way that we operated and it was a shift in our technology. You know, the way that we use technology is driven by the employee base. Your company's completely 100% in the AI generation with— and I've worked with your team. Sharp, sharp, sharp people. You know, everything is, you know, every— everything is clearly, you know, is fully AI-enabled when you're working with your team. So what's it like building a company today in not just the digital native generation, but now in the AI generation? Who do you look for? What makes it— what makes a great team player? And what are some of the things that you just don't have to deal with? Anymore when you're building a company in this way?

Yeah. Well, I think on the flip side, maybe one thing I'll say is that there's still a lot of things that I think are similar. I think, you know, especially as we're starting to scale more and more, you know, our customers say this and I truly believe it as well. At the end of the day, it's a people's business, right? Like every deal, like when we go and make deals and work with people, we don't ever really think of ourselves as just giving them a piece of software. AI, we really think about ourselves as being an extension of their team and we have to earn their trust every time. And so I think still the trust, the brand, a lot of the core parts of building great business like the one you've built, it's still, it's still very people heavy and that hasn't changed at all. I think on the AI native front though, that I mean, it is, it is definitely something that and even compared like compared to my last job in neuro, it's like so much is different. And I think the most interesting thing is that the difference between people— like before, you know, there was a Silicon Valley term that has got pretty popular as like 10x engineer. Like there's an engineer that can do 10 times as much as an average engineer. And I always thought that was like a bit of an exaggeration. It's like you're not really doing 10 times as much. You might be doing maybe 2 times, maybe 3 times if you're really good. But now with the new tools, not only engineers, but even people on the go-to-market team, in biz ops, in rev ops, like the delta between someone that is truly AI-pilled and is like using everything, agents, versus someone that's just slightly like, you know, uses like ChatGPT sometimes and like isn't as proficient, that that delta is being magnified. And so on the engineering front, I think an example I always like to give is we have a guy on our team who like up until recently was basically an IC, but he was genuinely getting the work of like 20 folks done because he actually had like at any given point in time, it was funny, he'd just be sitting. I was like, dude, like, what are you doing? Like, are you going to Are you gonna get some work done today? And he's like, calm down, I have 20 agents working for me. And these 20 agents are being managed by 2 AI managers that I've created that is reviewing the PRs. And then I only come in when like one of these things are ready for my review, and then I quickly process it and give it some feedback. And I was like, that is, that is miraculous. Like you're, you're effectively having like an entire, you know, digital team around you and it's like, and we do enablement and stuff, but I would say in this day and age, it's so important to find people that are just genuinely curious and are at the cutting edge. And even another example is our VP of Sales, Jack. This guy is incredible. Like he's obviously leading the sales team, but in his spare time every weekend, he's cloud coding some tool that is like incredibly useful. And even analysis, like I have an interesting analysis like, hey, you know, what was our, you know, NDR for this amount of time? Or like, what is our, you know, what do we need to do to hit this revenue goal in Canada by this time period? And I'm like, that's something that before would probably take at least, you know, a few hours, maybe even a couple of days. 3 minutes later, he has everything set up that he can give me a detailed answer. And it's really incredible.

So outside of Avoca, since you kind of gave a couple examples about AI use, what do you generally— what do you see? Is there anything that you think about or you're excited about, even though your head's down in Avoca, but as you look in other industries, robotics, things like that, self-driving cars that you have experience with, what kind of AI-powered enhancements do you see coming down the road that you're excited about or interested in following?

Yeah, robotics is always something that's so difficult because it relies on the hardware. I think that's why it took, it just, I mean, it's not even proliferated, but, you know, self-driving cars is a canonical example for The last 20 years, like, self-driving was always 5 years away, right? And it's now like somewhat in mainstream. Like, Tesla obviously has an incredible FSD system. My dad, he got one like end of last year, and he'll basically— he might intervene like once in like 10 trips. So it's really gotten really good. Obviously, Waymos are doing really well in in San Francisco and other parts. But I think robotics generally as a whole is always difficult. With that being said, I do think that, you know, some of the inventions, especially around, you know, household robots is something that piques my interest. I've invested in a couple of those. The other area that I think is particularly interesting is actually in drug discovery. And that's something I actually had done a bit of research myself with, like the early version of GANs, generative AI models, you know, back in senior, junior and senior year at MIT. And it never, it just never really got there. The tech wasn't good enough. But now I think there's actually pretty significant advancements and Uh, yeah, it's like, it's pretty, uh, I think the ability to create many more of these synthetic, um, molecules with AI that are able to just, you know, cure a lot of the biggest diseases. I think that's, uh, yeah, that's quite fascinating.

So are you talking about robotic molecules?

Are you talking about engineering Oh, no, I'm just talking robotics is a separate thing, but I was saying engineering new molecules for drug discovery. Yeah.

Nice.

Nice. Yeah, I was, I forget who it was I was listening to or watching a podcast, but they were talking about the constraints of robotics is largely because of the, you've got to have parameters. You know, you got to put a fence around it. It's got to know that it's going to operate within this area. So if you go to an Amazon factory, right, it's, it's within a variety. It's got the constraints as to what it can do and where it can go, and it can perform that work very well. When you talk about a humanoid robot, which are being produced now, many, many companies— Elon has sunset his Model X and his Model S and converting those factories to robotic production. You know, only with AI can they maybe operate without parameters, like, because they have to process so much information about what the input that's going on out there in the world. You know, so if they don't have a fence, like I was in a robotic mowing business years ago and, you know, we literally had to put a wire around the yard because it was, if it would just go everywhere, it would mow the entire neighborhood if you, if you didn't put a wire around it. And so yeah, it's, it's interesting to see how, you know, how those constraints are gonna be lifted. And then, you know, it was interesting to hear about the robotics You know, it's, it's when you really, when you really look at the human form, you know, that it's not, it's, it's the reason that they're building these in a humanoid form is because it's very efficient, right? I mean, you've gotta have, for balance, you know, we, the, the way that we're constructed, you've gotta have visibility from a high point. So you would, you would have a head up at the top where you put the cameras and everything like that.

Mm-hmm.

You've gotta have opposing digits and you've gotta have, you know, you can, you can only do so many things with one hand, but you can do a, a huge amount of things with two hands. So at the end of the day, the robots are going to look a lot like us.

Yeah, I mean, and also just the affordances. I think I was, I was also listening to a podcast with, with Jensen and Lex, and they were talking about, it's like, well, you don't want the robot to just like, there's all these tools, right? And these tools are made for humans and you need to like it has to be somewhat of a human form factor because you don't need like to heat up a piece of water. You don't need the robot to shoot laser beams and heat it up that way. You can just use microwave, right? In order to use the microwave, the microwave is built for humans and so is the washing machine and everything else, remote controls. So I think, yeah, it is certainly, yeah, certainly you need the, to some degree, the the form factor there. And, but yeah, I, I don't have a house yet, but I think hopefully when I have a house, the humanoids will have been, you know, more developed by then.

Yeah. Well, you're in New York. Houses are in short supply.

Right, right.

Exactly.

Yeah. Well, I think, but if you, if you, if you play that into the future, the way that homes are built, where the services are and all that, could be constructed differently so that specific for the use of robots in maintaining the home, right? So exactly. You could put everything below a certain level because the humans aren't gonna have to touch it anymore. Yeah. Light switch, electrical plumbing, you know, and I, I think, you know, we'll, we will shape our built environment around the, the services that the robots can do because all we're gonna be doing is, is just living in it. It's their world and we're just living in it.

Yep, exactly. No, I totally agree. I think, yeah, there's several startups also that are doing really cool, like kind of a similar concept, but for the internet, right? Like the internet is built for us, but in 10, 20 years, like most of the traffic in the internet is going to be agents. And agents don't— and agents obviously can just, you know, access APIs and access things much differently than humans. And so this company, this other company is called Parallel, is very interesting. They're doing, they're trying to rebuild the internet for AI agents, which I think is very smart.

Yeah. Yeah. Look, if every car was self-driving, would there ever be an accident?

Yeah, that's a good, yeah, I think that's a very good point. And yeah, I think if tested properly, probably not many.

Yeah, we won't be allowed to drive anymore, Tyson. Well, we can all live in New York. We won't have a car.

There you go.

There you go.

Awesome.

Well, look, I want to spend the time that we have left really digging into the how you're, how you're enabling the home service space. And, you know, just, I mean, the, our telephones and our websites, it's the front door of our business. It's where we meet our customers. It's really the most important interaction that we have. Is the first one and to be able to make a great first impression. So you've— are you completely specialized in the home service space now?

Yes, that is our number one thing from, from the, from the early days. We have— we were very focused in initial days serving HVAC, plumbing, and electrical and just, you know, really honing in, building the perfect product for them. Uh, and now I assume in the last year we've expanded our aperture a bit to include other, um, resi home services like, um, uh, you know, roofing, garage doors, pest control, restoration, um, and even like home energy and stuff. Uh, but our core focus still remains basically everything that touches a home and the service component of it.

Yeah. So when you, when you look at a new vertical, yes, it's in the home services. So you say, all right, we're not active in roofing, we're going to go to roofing right now. How much, how much work is there to understand that customer journey, the sales cycle? I mean, is it something that is, is, is months or is it, is it, is it a year thing to figure out, you know, how to really bring on a new customer type?

Yeah. Yeah, it's a great question. I think for us, it's months, if not weeks. I mean, we—

there's—

and don't get me wrong, there's a lot to understand, but typically our model is, you know, we think of things in concentric circles. So the first tier is like, you know, everything that's Resi Home Services. Obviously, there's roofing, there's fundamentally different workflows. It's a lot more outbound-based. It could be more— some folks do more storm chasing. But it's a lot more kind of outbound based than HVAC, where a lot— it's a lot more inbound. And so typically our model is we take a new vertical, let's call it roofing. We usually go in and we partner with a few design partners. Now it's very important to pick the design partners properly because you have some people that like, they're, they're, they're like, they just want to see results. That's not necessarily a good design partner. You want a design partner to be what I call early adopter, someone that's like, you know, genuinely enjoys the technology, wants to be an early mover, understands that the product initially isn't going to be perfect, but is willing to work with a company and, you know, get all the benefits of helping in the product development journey. And And so basically we get 3 of those, 2 or 3 of those, you know, we really iron out, build something somewhat custom at first. And then, you know, once they're really high NPS, really happy, then we start to figure out how to get kind of the rest of the industry. And usually we pick the large ones, so they're already influential. And so it's like, okay, if we already succeeded with these 3,, then the rest of the industry usually will follow suit. And that's the general strategy. But, you know, to your kind of question, it's for us, like, name of the game is speed. I think that's— you have to— the AI is moving so fast. And just generally, I think in any business, like the, you know, the velocity which you can ship and iterate is a huge indicator of how successful you can be. And so, you know, we think of things in sprints and weeks. If not days as we expand.

And when you interview a new client who's not using any technology, maybe they've got just a traditional call center, what are, what are some of the easy wins that you can point out to them that Avoca fills the gap and steps up and improves their business immediately?

Yeah, well, I always describe it as kind of multiple stages, but The first stage is just like really shoring up. A lot of folks, they don't realize this, but they're losing revenue left and right, and they're essentially, they're just leaking it, right? And so, you know, our initial wedge that we always started with was overflow and after-hours, because 9 times out of 10, folks do not have a good solution. It's either voicemail or it's some outsourced call center or some answering service that doesn't know anything about the business, isn't— doesn't integrate into the CRM. So it's essentially just a glorified voicemail. And so just by having an agent that answers on the first ring every time the way you want, that integrates into the CRM, just that alone, huge, huge uplift, massive uplift. Beyond there, there's all these other what I call holes that you can patch, right? How many, like Jeff, how many businesses, even like, you know, good-sized businesses, $10 million businesses, do you think are answering every lead in under 10 seconds and following up religiously? Or is taking every estimate and like hammering that estimate and reminding them, or when they're busy, when they're not busy, actually utilizing their membership base or previous customers to fill up their schedule religiously. Like, how many businesses do you think do that, like, even decently well?

Uh, not many.

Yeah.

So that's, that's kind of exactly the, the problem we, we saw.

Yeah, I think any, anytime I've had a vendor that comes in and says, look, I, we understand that you have this, this core capability, but if you just give us an opportunity to work around the edges and to You know, it's a clear win, right? People that, people that call in the evenings and the weekends, we always staffed in the evenings and the weekends because we knew that they have an incredibly high close rate. Nobody takes time out of their weekend to give you a call unless they really, really need something. And the same thing is in the evenings. You're, you know, if you're, if you're the one that's answering the call in the evening and the weekends and other people aren't, you're going to get that business hands down. So, so focusing on those areas. And then again, talk to me a little bit more in depth about mining for opportunity. Maybe when you're slow, how does Avoca help companies go back through their previous customers, maybe their, their debt estimates or their, their unclosed proposals and go back and help them convert those specific opportunities?

Yeah. So again, it's a little bit different based on the vertical, but, you know, maybe I'll just talk about HVAC, you know, a kind of broad, broad one there. It's like in HVAC, you have generally two really busy seasons, right? The winter and the summers. And then the other times, it's like, you know, you still have a huge technician supply, but there's not as much like, you know, big things to do. And so here, usually the concept is we actually look at the weather and we predict like how many demand jobs that you might have in any given week or any given day. And then we backfill all the rest with things like inspections or free maintenance visits, you know, if you're a member, um, or, or other things, um, too. Uh, and so, and the way we do that is we, we look at, you know, all of the folks in your, in your membership list, in your, in your CRM or whatever data source that has that. And we also look at any customer that has ever made contact with your business over the last 5 years. And we, and we make it really simple, right? We, we want to make it as simple as possible and as frictionless as possible for the customer to book. And so we just provide time slots. It's like, hey, would you want an inspection or free maintenance for your HVAC unit? Here's the times that are available. Just reply 1, 2, or 3. That works incredibly well. And, and then it's also flexible in the sense that once those days are filled, We stop sending out messages. And so you're able to both—

Is that a voice outbound call that you're making, or is that a text?

Sometimes, sometimes it's a voice, but, um, other times we just do text because that's— we find that to be actually very, very useful.

Oh, that would be— that would be fantastic if somebody reached out to me for a service that I wanted and I had just had to press a button and know that it was scheduled.

Yeah.

I mean, I think this is honestly— and obviously I talked about HVAC, but this is like really broadly applicable, right? Like, like, you know, my girlfriend's a dentist and this morning she had like 3 people cancel on her and it's like, well, now it's like, yeah, it's kind of unideal. Then, you know, you have a huge gap in your schedule. Imagine an AI that just figured out, okay, let's fill, let's fill it up with folks that haven't had a, you know, are overdue for their for their cleaners.

So there's like, I think, I think this general concept of just identically like finding holes and plugging them with any service business that has like a schedule or dispatch board is pretty powerful.

Yeah, it's an optimization play.

Yeah.

Well, Tyson, What should I have asked you that I haven't yet?

Yeah, I mean, I guess I can—

what's exciting, like what's exciting about Avoca into the future here? I don't have the numbers exactly, but I know that you guys closed on a round of funding.

Yeah, we did our Series B, actually it was end of last year. We just announced it a few months ago. And so, yeah, that's been, that's been great. Brought on some really good partners and obviously secured a great amount of funding for all the R&D and efforts that we're doing. But I think, yeah, maybe I can spend a little bit of time touching upon, you know, what gets me excited, the vision of the company.

Yeah, that'd be great.

Yeah. And so I think, yeah, like, you know, obviously we talked in depth about some of these things. How some of these tactical things and kind of the origin story. I'm really, I think the thing that gets me most excited and out of bed, you know, pumped every day is really the impact that we can have bringing AI to the entire physical economy. And so, you know, I think there's so many companies that could benefit and traditionally AI was, was really only accessible to Fortune 500 companies, right? That's kind of the, those large-scale implementations I did early in my career. That was like the only thing. But I think where AI can actually have the most direct value add is actually in the, in the blue collar, in the physical services. And so that's really what we're excited about. We want to give every one of these companies a, an AI agent experience that's Fantastic. I can really uplevel them. And it's not about replacing people. I think that's a very common misconception narrative. Obviously, there will be some people that will have different jobs, but I think it's really about, uh, you know, allowing and giving, giving companies, uh, SMB companies, you know, folks in the physical trades, the ability to scale a lot more aggressively and quickly than they were. Like, our dream is to be able to make it so that a 20-person company, 100-person company, instead of taking 10 years to go from $1 to $100 million in revenue, we want them to be able to go $1 to $100 million in 2 or 3 years. And we want to— we just want to empower everyone with the tools and means to do that.

Yeah, that's incredible. What in terms of, you know, future raises, things like that, anything you can share there? Or is this a public company at one point? Can't really say, probably.

Yeah, the goal is to go public one day. I can't share too much on future raises, but I can say, yeah, I mean, we've just— did the Series B not too long ago and, and fortunate to have not even really touched that at all yet. So I think we're really excited. We want to grow fast, but also sustainably. And yeah, I think there's just, there's just so much opportunity in the space that we're both in. And I think for us, it's, it's really about, you know, empowering these folks and, and creating something that can that can be extremely value-additive to all the folks out there. And I think one of my favorite examples is it's like, you know, during these large snowstorms or heat waves, oftentimes, you know, people aren't even able to go into their shops because there's like a huge outage. Phone lines don't even ring, and Avocate can still be on and allowing our customers to be able to serve their communities. That's what really gets me excited.

Fantastic. All right. I've got a curveball and a fastball for you to wrap it up here today. But before we do, tell people how to get in touch with either you or Avoca. Where do you want to direct people?

Yeah, I'm always, I'm always comfortable sharing my direct contact. It's just tyson@avoca.ai if anyone ever wants to get in touch. And then if you're interested in learning more website, you can see a few demos and this and that and obviously book a demo if you're interested. And that's just avoka.ai. Awesome.

All right, here's the curveball. Gun to your head, you've got to start a business in the next 30 days. It can't be something that you're currently doing or currently active in. Where is the opportunity in the market? Where would you start that business? You have to start one.

That's a good one. I think if I were to start a profitable business, um, I would actually choose I'll give you two, one of two things. I'd choose an AI consulting operation because I think there's just a huge opportunity. I know some of my friends have kind of taken that leap and yeah, I think there's just a huge opportunity to actually bring AI into businesses. And so I think that'd be something that I'd be pretty good at. Or I would actually I think with all the things that I know now, I would actually start a home service business and maybe actually franchise one of your shops, Jeff.

So, well, here we, you know, we great 4,500 great markets still available, Tyson. Yeah.

You know, I think that that would be my second choice.

All right. I agree. I agree. Home, home, it's very durable. They're essential services. They're not going anywhere. There's no obsolescence to home and property services the way that we do 'em and, and and the way that we communicate with our customers, the way that we route our trucks, everything is going to change for the better for the customer because they will get better service in a more timely manner. And of course, we'll be more profitable and we will, we will prosper. So with help from you. But all right, last question, Tyson, fastball straight down the middle. You've got one sentence to make a difference in somebody's life. What is it?

One sentence. Always strive to, uh, always strive to be 1% better every single day.

Perfectly said. Tyson, thanks for being on.

Thank you, Jeff. It was a pleasure.

Yep. Awesome. I'm Jeff Dudan. We're here with the incredible Tyson Chen with Avoca, and we have been on the Unemployable Podcast. Thanks for listening.