Only Book What You’re Willing to Own

How AMII Helps Businesses Adapt to AI with Adam Danyleyko

Adam Danyleyko: [00:00:00] We work with hundreds of startups a year, helping them see where AI can be applied to their business. A lot of those startups are coming to us. Not knowing where to start, even they're like, they don't know where AI can be applied to their business. They don't know what, how it can look or what the different options are, and we're helping them identify those opportunities within their own business.

Intro: The Great Mark Cuban once said, business happens over years and years. Value is measured in the total upside of a business relationship, not by how much you squeezed out in any one deal. And we couldn't agree more. This is the Business Development Podcast. Based in Edmonton, Alberta, Canada. And broadcasting to the world, you'll get expert business development advice, tips, and experiences, and you'll hear interviews with business owners, CEOs, and business development reps.

You'll get actionable advice on how to grow business, brought to you by [00:01:00] Capital Business Development CapitalBD.ca. Let's do it.

Welcome to the The Business Development Podcast, and now your expert host, Kelly Kennedy.

Kelly Kennedy: Hello, welcome to episode 314 of the Business Development Podcast and today. It is my absolute pleasure to bring you Adam Danyleyko.

Adam is a dynamic leader and passionate advocate for the growth and success of startups in Alberta and beyond. As the product owner of startups at a Adam leads the charge in helping early stage companies explore, build, and deploy machine learning and AI technologies with a rich background that spans leadership roles at Startup Edmonton, and over four years in strategic positions with the government of Alberta, Adam's experience positions him at the forefront of fostering innovation and growth in the tech ecosystem.[00:02:00]

His leadership has directly impacted dozens of startups scaling their ML adoption, and ensuring they have the tools and support necessary to succeed, driven by passion for community engagement and technological innovation. Adam's impact goes beyond the office. His expertise in AI and machine learning has empowered countless startups to unlock the potential of these technologies, helping them scale and transform industries with a sharp focus on guiding companies.

Through the complexities of ML adoption, Adam ensures that each startup is equipped to leverage cutting edge solutions that drive real world impact. If you are ready to take your startup to the next level with ai, Adam is the leader who will help you harness its full potential. Adam, once again, it's a pleasure to have you on the Business Development Podcast.

Adam Danyleyko: Great to be here. Thanks so much for having me again.

Kelly Kennedy: This was Mortal Kombat. We would be yelling round two fight. [00:03:00]

Adam Danyleyko: I love it.

Kelly Kennedy: Yeah, man, I, I owe you such a big apology. Uh, the first time we did this show, guys was literally recorded probably like four months ago.

Adam Danyleyko: Yeah.

Kelly Kennedy: I, I dunno what the hell I was thinking that day, but somehow my audio, like my levels got put too high and I sounded like I was underwater.

I completely blew out my entire audio track. Adam sounded incredible and I sounded silly.

Adam Danyleyko: Well, I'm glad we're able to, uh, to finally get back in the studio and chat again.

Kelly Kennedy: I'm really happy to have you back, man, and I'm really excited for this conversation because the first time was really epic. So the second time it's gonna be even more.

Adam Danyleyko: Oh, well, I hope I can live up to the expectations of, uh, of our first conversation.

Kelly Kennedy: No kidding, no kidding. Well, this is really important to me because, you know, when I started the business development podcast, one of the things I really wanted to do, we shine a big, bright light on Alberta and Edmonton in particular.

Yes. And Calgary doing some pretty incredible things. But I wanted to do something called the Alberta Ecosystem Series, because what I realized pretty early on doing this [00:04:00] show was the startup ecosystem in, in Alberta's. Pretty incredible. But like, the amount of things going on is almost like hard to fathom.

There's just so many different things going on. Whether you're talking about, you know, Alberta Innovates, PrairiesCan, you know Edmonton Global, Edmonton Unlimited, Startup TNT, there's BDC Canada, there's like a million things out here to help us, which is great, but it's really hard to understand the whole thing.

Yeah, and I remember hearing about AMII pretty early on, that you guys were doing some pretty incredible things with, with ai, but I was like, oh my gosh. Okay, one more thing we gotta just add to the list, because it's like hard to understand how it all fits together. So we are going to do that today. By the end of this show, people are gonna understand what AMII is and how the heck they can utilize it.

But before we do that, dude, you work in one of the coolest tech sector industries. There is. How did you end up on this path, man? Like, who is Adam Danyleyko?

Adam Danyleyko: Yeah, it's, uh, it's kind of wild to think about the fact that this is where I've, where I've ended [00:05:00] up. I actually have a, uh. My bachelor's degree is in human resources from the U of A, uh, here in Edmonton.

So, started out with in human resources, like you mentioned, I, I worked at the government, uh, of Alberta here for a few years in a variety of different roles. I started in hr, so I was actually HR business partner, and then I moved into some policy work supporting a climate technology task force. Ran a cross-government internship program for the government.

Kinda my last role there. That's kind of how I got into kinda the student engagement side of, uh, my work. And then that's, that's what led me to Startup Edmonton. So from there I moved over to Startup Edmonton, where I led out their student engagement. Work and I built a team around that. And, uh, from startup Edmonton that's really how I got into the tech ecosystem, right?

Like meeting startup founders across the city, really, um, being engaged with that community and really fell in love with the community there, seeing everything that, that these amazing founders, um, and other you developers are building here in the city of Edmonton. And then the opportunity, this [00:06:00] opportunity at AMII came up, uh, to come over and build up the startup team here at AMII.

Uh, so being able to come and leverage my experience, uh, with the startup community in Edmonton to know what founders needed and what kind of supports were, were required for founders to come here and build out, to look at the a e programming and see, you know, what was working for startups, what wasn't working.

Yeah. What, what could be modified? What, what do we have to build from scratch in order to see how can we support startup founders? Not just here in Edmonton now, but across in Edmonton, across the, uh, across the province of Alberta, uh, across Canada and internationally. See them. How can we support startup founders better along their AI journey?

Um, and so that's kind of, Cole's notes journey of how I got, uh, into my, into my role here.

Kelly Kennedy: It's super, super cool. And, you know, I feel like on a certain level, and I know this isn't true, but to me it felt like AI just came out of nowhere, man. And you know what? Like, I don't know, maybe I just didn't have my ear to the ground.

Maybe I just wasn't paying attention. Right. But it [00:07:00] felt like 2023 hit and there was like before chat, GPT and then after chat, GPT, like, you'll literally, someday someone will draw a giant line in the sand and say, this is where everything changed, right?

Adam Danyleyko: Yeah. Oh, I, I think that's super fair to, uh, you know, for the average person, I think that that's accurate.

It's, I think when we started seeing chat, GBT was probably one of the first tools that that hit the, like the. Everyday person and, and, and, and it was also doing something that for a long time seemed to be well within only the realm of humans. Yeah. Um, and it was now putting out text and, and words and it was responding to questions in a way that, you know, we, there's been chat bots for a long time.

They haven't been very good. Right. It's been like asking very specific question. You get a, a pre-generated answer back and it was just kind of key words back and forth. Now you're getting. Answers back when you ask it a question is able to do a lot more than just answer your, your standard 10 AQ questions.

Yeah. Um, and so I think that's, that's pretty fair. I [00:08:00] think though, that, what's interesting is from AMII's perspective, we've been doing this for over 20 years, coming up on 30 years now. The province of Alberta invested heavily into AI back in the nineties. And now we're, you know, that's what led to the creation of AMII.

Um, and where we're at today is it's, it's been around and it's been something that the U of A has been doing for 20 to 30 years.

Kelly Kennedy: Wow.

Adam Danyleyko: And like, if you think about like, AI as a field stems back to the fifties Yeah. Computer science goes back even farther. So this is a field of study that's been around for a long time, but now it's finally breaking into kind of the, the every day of, of most people's lives in a way that you can see and you can interact with.

Yeah. That's obviously ai. Um. Even, you know, AI and machine learning has been a part of our lives for a long time. Yeah. We just didn't recognize that that's what it was. Or, you know, the average person maybe wasn't recognizing that they're interacting with ai. Whereas when you see Chat GPT, it's like, it's obvious that this is, is an AI that I'm [00:09:00] interacting with.

Kelly Kennedy: Yeah. Yeah. It's crazy. It's crazy because like, dude, even when I started this show, like the very first episode, I think I actually learned about Chat GPT only about 20 episodes into the business development podcast. So it's like literally I've been using AI for less or for for less time than I've been doing podcasting for, which is a little bit like unbelievable because now we utilize it so heavy in literally everything from like idea planning to post creation to content creation, to running ideas to fact checking, like.

There's a thousand things that we use AI for that like in the very beginning I didn't use it at all. So it does feel like 2023 was just this like light switch. It's like, oh, it's here. Use it.

Adam Danyleyko: Totally. And it's like, it'd be interesting to hear your perspective on that because it's like, were those tasks that you're now using AI for tasks that you were doing a different way previously?

Yeah. Or were they things that you just weren't doing at all [00:10:00] or both?

Kelly Kennedy: Yeah, I would say both actually. So they're like, honestly, what AI has enabled me to do is essentially run a podcast on my own, right? Like, I can do a full show production. What, what, what would've required an entire team before? And you know what?

I've had to tweak it along the way. I've had to find ways to insert myself back into it because I don't like to use completely AI generated content. I like to like have some Kelly in there too. Yeah.

The

cool thing about Chat GPT is it learns who Kelly is too, right?

Adam Danyleyko: Yeah.

Kelly Kennedy: I feed enough of my information into it and say, okay, this is me.

This is how I write, this is how I talk. It's pretty, pretty incredible what it'll come up with just from that data.

Adam Danyleyko: Yeah, a hundred percent.

Kelly Kennedy: But yeah, like in the beginning, absolutely. I was writing everything. I was writing my episode descriptions, I was writing all the, you know, like everything.

Everything. Yeah. Like I can't believe like how much time savings AI has really put back in my pocket. Not just time savings, but like level up savings, right? Like, it's not just time savings. It's like a massive level up in my ability to perform work. And you know, I'm, I'm one example, right? I'm a [00:11:00] podcast, I'm one example, but I can't imagine what if you scale that, that out, how big of an impact that is gonna have on enterprise and business as a whole around the world.

Adam Danyleyko: Oh, for sure. And I think that's kind of where we see AI working the best is like in those situations where it can make humans better at what they do. And that's kind of the sweet spot I think in a lot of ways is how can we level people up? How can we take people that are already doing something and make them more efficient with their time?

Yeah. Take away those kind of tasks that they don't like doing in the first place, or might take up a lot of their time, might be boring or repetitive but might also at the same time require like a lot of accuracy. Yeah. That's a perfect place where AI can come in, is like, it's gonna be accurate. Like it doesn't get tired, it doesn't need a coffee break, it doesn't need to go to the bathroom.

Kelly Kennedy: Yeah.

Adam Danyleyko: Like, so it's not having those kind of human ebbs and flows of attention yet at the same time it's able, and it's able to work all the time. So it's able to maintain that consistent level of [00:12:00] output. Which I think is, is, is really helpful to reduce those kind of menial parts of a lot of people's jobs, um, and level them up in a way that allows 'em to be more effective, more strategic, more impactful with it.

Kelly Kennedy: Yeah. Well, and I would say too, like the upgrades that I'm noticing with Chat GPT are like, they're, they're pretty massive man. Like, they aren't just like little leaps as they upgrade these models when they put out a new model, like 4.0. Like for instance,

this is coming out in February of 2026. So like, who knows where we're at, at you a year from today, but I know as of right now with, with Model 4o, with Chat GPT, it's pretty frigging incredible. Like, I find myself a nightmare sometimes just having like full on conversations with like, it's crazy.

It's crazy though. That's where we've, but like, it's fun. It's fun to have those conversations with Chat GPT.

Adam Danyleyko: Yeah, it's interesting. It's, yeah, it's pretty smart. And it's, it's really cool how it works kind of under the hood and, and the fact you can have like a real proper conversation with something [00:13:00] that you know, doesn't.

Doesn't really exist in a weird way, like it exists, but it's like, like you can have a conversation with it, like you're talking to a person, but it's, it's not a person.

Kelly Kennedy: It's, yeah, It's crazy though because like, I'll just be like bouncing business ideas off of it. I'll be bouncing like, naming ideas or like show ideas or topics or like just asking you to search shit on the internet.

And it's like, it's pretty frigging incredible. And I think about the fact that like, at this point, we're talking maybe two years later to the very first time that I experienced utilizing an at, you know, a, a machine learning or a, or a Chat GPT to where we are today. And I just think, holy crap man, where are we gonna be five years from now?

Unreal. Yeah. Unreal.

Adam Danyleyko: Totally. Yeah. No, I agree. It's, it feels like it's, it's changing exponentially. Um, and, and you know, even working in the field, working with scientists every single day, working on projects with clients every single day. It's still hard to keep up. Wow.

Kelly Kennedy: Yeah, I bet, I bet. So take me into AMII.[00:14:00]

You know, you mentioned that, that AMII's been doing this stuff for 20 years. Like, my gosh, my head has just been in the sand. Like, how did you guys keep this stuff under wraps for that long?

Adam Danyleyko: Yeah, so that's a good question. So I guess, AMII, as we currently exist, it's only been around since about 2018. Okay. Um, but.

The U of A's AI department, um, has been rebooking into AI machine learning for that kind of 20 to 25 to 30 year mark. And so it's, it's really been on the research side for a majority of that, of those 30 years. And still a majority of what we do is on the research side. Wow. So we have research chairs and fellows that, that we fund at the University of Alberta.

These are world leading researchers, like, like top of the world researchers in ai machine learning. The U of A has actually historically been ranked in the top five, you know, it's number three. Number four for AI and machine learning research internationally of all universities in the world. Wow.

And so it's, uh. It's a real foundation of what we do Here is the research side, and it still [00:15:00] is to this day. So, we are actually in the middle of growing our researcher tool. I think we're around, you know, by the time this podcast comes out, it might even be up to like 60 researchers. Wow. Um, cross campus researchers studying ai, machine learning, everything from applied to really theoretical research as well.

And Dr. Richard Sutton is our chief science advisor and was one of the first run researchers that was recruited to the U of A during that time. Yeah. Um, and he is one, he is the, the kind of the founder that's kind of, he is the founder of Reinforcement Learning, one of the three major studies of machine learning.

And he literally wrote the textbook on it. And we have him at the U of A. He's at AMII, um, he's our advisor. So, you know, we're very lucky to have someone like him, as a part of our organization. Uh, and, and then all the researchers that have come with him and come after, uh, that as well that really built out that side of our business.

And then now we're at the not-for-profit. We're able to leverage all this research, all these, this [00:16:00] community of students and researchers, uh, that we have to support industry. And so that's really what, what AMII as a organization does, is we fund that research and then we work with industry to help translate that research into industry adoption.

So we work with everything from startups to mom and pops to, you know, national corporations not-for-profits, multinationals help them along their AI journey. Everything from training, like what is AI machine learning at kind of that, that foundational level, that basic, um, knowledge and building a common nomenclature at your organization around AI all the way through to advanced certificates in ai, machine learning.

We can help companies. Identify where AI can be applied to their business. Yeah. Scope that out to build that roadmap on how can you actually go about building and implementing that solution to doing the hands-on ml proof of concept development for our clients. Um, we support along that whole journey.

So it's, um, it's a lot of fun. There was a lot of things happening and kind of [00:17:00] going on all at once. But that's kind of maybe like the, the super quick Kohl's nose having to dive in on any of those other pieces as well. But that's kind of the, the really high level idea of what we do. But maybe one, one more thing I'd say is, I think something that's really, really good about AMII and something that, that really.

Aligns with me personally is, is our, our mission is AI for good and for all. Mm-hmm. So it's, and, and that's really at the core of everything that we do, is that for good and for all, um, when we're bringing on clients, we actually put them through a principled AI rubric that's based on the un sustainable development goals to see is this a project that we can and should be working on?

Interesting. So it's, we take it very seriously. And we're very conscious of the impact that any, any of our work will have on people on the planet. Yeah. Um, on everything that, that we, that all the important things like we're very c conscious of bias and insights and fairness and, and the [00:18:00] impact of our work as it relates to everything that we do as well.

Kelly Kennedy: Well, it's very interesting because I don't think many people are, are thinking about how powerful AI is. Right? Like, it's amazing, and, and I think, I think you guys are doing it right by starting out with a set of ethical guidelines to work from. You're kind of setting the tone for how we're going to approach AI moving forward.

And I think that you're absolutely right. That's super important because when you have this much knowledge in one place, right? Knowledge can be used for whatever you wanna use it for good, bad, or ugly.

Adam Danyleyko: Yeah. Oh, for sure.

Kelly Kennedy: It's, uh, it's interesting, I, I'm sure not since, not since the advent of nuclear energy, has somebody had to create an ethics guide for how we were gonna utilize a tool.

Adam Danyleyko: Yeah. I mean, I, I, I think that's really fair. I think that's probably as you know, an interesting parallel for sure. But I think at the end of the day, like ev ai, like you said, it's a tool. AI is a tool. Building tools is what we do as humans. Yeah. [00:19:00] Um, and every tool can be used for a multitude of, of applications.

Uh, like you could use a hammer to hammer a nail or to break a window. Yeah. The hammer is just a hammer, but this is how you use it. And I think, uh, we, it's, it's really important though within that to have that ethical framework. And like I said, we call it our principled AI framework in order to make sure that you're taking into consideration the impact of the tool that you're building.

And using we, it applies to both how we use AI internally. Yeah. Um, but also how we build the AI that, that we do for our clients.

Kelly Kennedy: Amazing, amazing, lead us into machine learning, right? Most of us just see chat, GPT, and we ask a question and we're like, how in the world is this thing doing this?

You know, for a lot of our listeners, AI is like, magic me included, man, right? Like, it looks like magic. It's like, how does this thing know to answer this so accurately for me and create these things, you know, I, I know you're not the technical scientist on this, but yeah. You know, please [00:20:00] do your best.

What is ai? What truly is machine learning? What's happening?

Adam Danyleyko: Yeah, so I mean, I think when you think about it, kind of taking even a step back, computer science is a field of study that's been around since like the 18 hundreds maybe even earlier. I dunno, some monks started thinking about this like hundreds of years ago.

Uh, and then within computer science, there's ai, artificial intelligence, and that is the study of making computers. Replicate the way humans think. Okay. Is essentially what it's often using algorithms or models might be different ways to think about that. But essentially, an algorithm is a math equation that is trying to mimic the way a human thinks.

Okay. Um, machine learning is a subset of artificial intelligence. And machine learning is building computer systems that learn over time and get better and improve, uh, over time. You know, some, some really common like ways that, like applications, like I guess within machine learning there's three major areas.

There's supervised [00:21:00] learning, unsupervised learning, and reinforcement learning. Those are the three areas. Like I mentioned, Dr. Rich Sutton at AMII was kind of the, the pioneer of reinforcement learning. Yeah. Um, which is where. This system or an agent is in a dynamic environment and makes decisions and then learns from those decisions.

So that like often you might see like an agent playing a video game and it either you put it in the video game, it doesn't know the rules, it takes actions, it gets rewards or punishments, and then it learns solely over time how to play the game. Yeah. Kinda like it in a lot of ways people like to think about that as like, how would a child learn?

It's, you know, learning by trying and getting feedback from the world.

Kelly Kennedy: Yeah.

Adam Danyleyko: Um, supervised learning is you have data, it has labels. You train your system on that. And then, so for example, you have a bunch of pictures of cats. You have a bunch of pictures of dogs. You tell the system, this is a cat, this is a dog.

And then you give it a new picture and it has to say, is this a cat or a dog? Yeah. Um, that's a very simple version of it, but it's essentially training based on data [00:22:00] that's labeled. And then unsupervised learning is where you. Give the system data, there's no labels and it, it is either looking for outliers or clusters.

And so this is, and a common application of this is your Netflix algorithm where it's recommending things to you. It's saying, Hey, based on you and everything you've watched, you are similar to these other people and they liked, they also watched this show that you haven't watched. So because you're similar to these people, you'll probably like this show.

Yeah. Um, so clustering or outliers, which is like fraud detection in credit cards, it's fraud detection and saying, I don't know anything about any of this, but this looks different from all the other ones. So it might, that might be, but that's kind of machine learning. At a, at a, the most basic.

It's like, and then over time, the more you train, the more different examples the system is seeing, the better it's gonna get over time. And the accuracy, speed, performance are all gonna be things that we're looking to.

Kelly Kennedy: It's incredible when you think about it. Like [00:23:00] I, I honestly can't fathom how much computing power has to be going in on the background simply to answer one of my random Chat GPT questions, right?

Like, I could ask just about anything on planet Earth and chat. GPT is going to search the internet and find me an answer and not just find me an answer, but write it to me in exactly the way I need it to write it to me based on the way that I asked the question.

How is this even possible? How big, like, is there like a gigantic computer the size of a city somewhere processing this information?

Like how is this, I mean

Adam Danyleyko: Yeah, kind of like, I mean, so data centers, there's data centers all over the world that, um, that, that have been being built for a long time. So when you think about the, the cloud, it's just data centers in different places around the world. So Google, Amazon, Microsoft. There's countless different other companies that, that all manage data centers that are used for not only storage but for computing, uh, [00:24:00] capabilities as well.

Uh, so yeah, like maybe not all in one city, like a city size block, but it's like a distributed network around the world of essentially supercomputers, uh, that are all doing these calculations. Yeah. And supporting it around the world and it's yeah, like the large language model type system. So I guess like, you know, I, I described three discrete types of, of machine learning.

Large language models are using a combination of all three of them. Okay. Uh, within that in order to actually do what they're doing. Uh, but they've ingested like chat. GPT essentially is trained on the internet. Okay. So it's like, it know, like it's, it's trained on all this data about everything. And what's really interesting is Chat GPT doesn't really know anything like it sounds like in knows things when you're talking to it.

Essentially what it's doing is it's predicting the most likely thing to say back to you based on what you said. So it doesn't like when you ask it like, you know, how [00:25:00] do I don't know, like how many cows are there in the world? Like it doesn't know what a cow is. It's, it's at the end of the day, like it's all using ones and zeros in a very fundamental way.

Like if you wanna break this super down and it's predicting what the most likely next character in the sentence is gonna be, wow. Um, the next word is gonna be, but it's very, very, very powerful 'cause it's been trained on everything. So it's using the context of what you're asking is using context of everything that it's been trained on to make those very, pointed, uh, predictions, um, within that.

Kelly Kennedy: Wow. So it, it actually is, is somewhat guessing, but it's guessing based on like a really educated experience, if that makes

Adam Danyleyko: sense.

A hundred percent. That's exactly, yeah. Like that's all machine learning at the end of the day is like, it's making predictions. Yeah. And so whether that's a prediction on will you like this movie based on the other movies you've watched, or will you buy this other item on Amazon because of everything else you bought?

Or is this [00:26:00] email spam or a promotion? Or should it go in your inbox? It's just a prediction. At the end of the day, that's, that's really all it is. And what we do with that prediction is where it becomes powerful, right? Like the prediction's great, but it's how do we as humans implement, uh, that prediction?

What kind of judgments, decisions, actions are we taking based on that prediction from this, from this AI system or machine learning system? That's kinda where, where the rubber's gonna hit the road. And then where it becomes useful, because a prediction for prediction sake doesn't mean anything.

It's how we, yeah. Use that prediction.

Kelly Kennedy: Let's just take Amazon for instance, right? Yeah. Whenever I buy something off Amazon, I immediately go down, before I check out, it says, people who bought this almost always buy these things with it. And dude, like nine times outta 10. I'm like, yeah, I need that. I need that.

It's so flipping helpful. Yeah. Like honestly as like, as scary as maybe some people might find like the idea of like ai, it's just so flippant, helpful. Like there's no two ways about it. Like Amazon, like, uh, you know, [00:27:00] predicting the things that you might need If you're doing, you know, a project, you might need that drill, well now you need drill bits, now you need some extra batteries, whatever.

Right? Like, it just knows what I need. And I would say most of the time I actually need that and I really do add them to my cart and buy them.

Adam Danyleyko: Yeah. What's crazy is like, I've actually heard I don't know, this was a long time ago, so I don't know how accurate this still is or, or if you know, it's something that they'd ever actually implement, but I heard at one point that Amazon was considering sending you things.

Just be like, Hey, we know you're gonna need this. Yeah. And then either you being like, I actually don't want it. Send it, but take it back. Or being like, oh, well it's already here. So I'll just say like, oh, like you bought a drill. So it's like they just send you drill bits. Yeah. Or they're just like, Hey, you're probably outta toilet paper.

So like, here's some toilet paper. And then if you accept it, then like, then they bill you for it.

Kelly Kennedy: Wow.

Adam Danyleyko: Because eventually, like they, theoretically, at a certain point in time, they'll be so good at knowing when and what you'll need. Yeah. That they could just send it to you knowing that you will need it by the time it gets to you.

That's a little dystopian in a way, but like it be pretty cool.

Kelly Kennedy: Dude, speaking of Amazon, [00:28:00] um, I was driving in LeDuc I wanna say like six or eight months ago.

Adam Danyleyko: Yeah.

Kelly Kennedy: And a, a drone, the size of a car flew over my head.

Adam Danyleyko: Oh, wow.

Kelly Kennedy: And turns out, I believe it's drone delivery, Canada's drone. And the thing is like the size of a car, man, like,

Adam Danyleyko: oh wow.

Kelly Kennedy: It blew me away. But obviously they're testing it around that airport area and I was just like, holy crap. This is the feature here. There are flying cars.

Adam Danyleyko: That's pretty cool. Yeah, I haven't seen that. That's awesome.

Kelly Kennedy: Yeah, man. But it won't be long in my mind until we have, if that exists, that only exists for a reason.

That reason is it's gonna be delivering shit from city to city. Right. Like if you think of a drone, the size of a car, its only purpose can either be to transport people or to transport things to another place. Right. So in my mind, my thought is, is like pretty soon you'll be able to order something, and Amazon will be able to ship it from Calgary to Edmonton on one of these little drone things in like no time flat.

And nothing will be on your door within like eight hours of you ordering it from like a [00:29:00] completely different city.

Adam Danyleyko: Yeah. Well, I mean, Amazon already in like major cities like New York and la they already have like 30 minutes delivery on a lot of items. Yeah, like 30 minutes you order it and it's in your house in half an hour.

Like, I mean, in Edmonton, I think the fastest you're probably gonna see is like maybe later that day or the next day, but like 30, like, I mean, if you have drones and like you have unfettered access, like I don't see why, why like you couldn't have things delivered within an hour or two for most items.

Kelly Kennedy: Me and you are probably pretty close to the same age. I, you know, I'm, I'm 36. How old are you?

Adam Danyleyko: Yeah, 32. I'll be 34 almost by the time this comes out.

Kelly Kennedy: Okay, awesome. I'll be 37 by the time. There you go. But you know, dude, we remember, you know, I mean, you're from the generation. You probably had N 64.

Did you have Sega?

Adam Danyleyko: No.

Kelly Kennedy: Okay.

Adam Danyleyko: I had friends that did though.

Kelly Kennedy: Okay. Okay. So Sega, like when, when I look back to being a kid, that was like the first video game system I had with Sega. Yeah. And I love that man. Like, I love playing Sega. And if I think of like, where we are today with regards to like [00:30:00] computer generation ai dude, it's hard to believe when you came from like 32 bit hay guy.

Adam Danyleyko: I know, right? Yeah. Well, that's crazy. Well, even just like, thinking back, it's like where, like, what was the world that our grandparents grew up in? Like, my mom grew up in a, on a farm that they had to stack hay bales around the house in the wintertime to keep it warm.

Kelly Kennedy: Wow.

Adam Danyleyko: Like, yeah, there's seven or seven of them.

Seven of them. And they lived in basically a two room house.

Kelly Kennedy: Yeah. Wow.

Adam Danyleyko: Like, so it's like, that's like, that's my mom. And to where like my kids, the world my kids are growing up in.

Kelly Kennedy: Yeah.

Adam Danyleyko: It's like, it's completely different.

Kelly Kennedy: Well, it's, it's weird. 'cause like I even look back to like, for instance, iPhone. Right?

Like, I was already out of high school. Yeah. When iPhone came out. I remember think at the time, I was like, oh, that's cool. But it's like kind of gimmicky and now it's changed absolutely everything. Right? Like you can't even buy a, a non-video phone anymore. Yeah. A non green phone, right? Like, it's just, it's crazy how much the world has changed.

And I just think sometimes, you know, my kids are really little right now my oldest is [00:31:00] 11 youngest just is little over one. And I just think of the world they're gonna grow up in, man. Like, I'm not even gonna be able to relate to that world. Like, it's just crazy how fast technology is advancing. And I think AI is only gonna put like a gigantic, a gigantic, you know, mock one speed to how that's gonna continue to evolve.

Adam Danyleyko: Yeah. Oh, of course it, uh, yeah, I, I have two little ones. I have a 2-year-old and a two month old right now. So it's, uh, yeah, it's, it's interesting. I remember when we got our first computer, like we had a computer in our house.

Kelly Kennedy: Yeah.

Adam Danyleyko: It was in the computer

Kelly Kennedy: room

Adam Danyleyko: and it loaded images about fast.

Kelly Kennedy: Yeah.

Adam Danyleyko: Yeah. Couldn't I only you, you couldn't use a phone while you were using it, right? Yes.

Kelly Kennedy: Yes. You remember a OL when they would send you CDs for their email internet?

Adam Danyleyko: Yeah.

Kelly Kennedy: Like, I just, it's crazy. It's crazy because obviously they were working on AI even at that time. [00:32:00] Like, that's hard to believe, isn't it?

Adam Danyleyko: Yeah. And, uh, yeah. I mean, it was, it, it was, it was very different AI than what's getting marked on today. But like, even like some of the, like there's lots of theoretical things that have been, conceptually figured out decades ago that are only starting to be able to be made possible because of.

Like the ability to, like the compute levels that are available today versus were available 20 years ago when those kind of theories were, were created or developed.

Kelly Kennedy: That's interesting. You know what, obviously computer technology and like, like you said, compute levels seem to be exponentially growing, um, exponential growing.

Do you think that this is gonna get even faster with the advent of ai?

Adam Danyleyko: I mean, I think so. I would say compute is what enables AI Okay. To exist. I mean, I guess like, in a way, like if you reverse engineer it, like you can potentially use AI to make better computers. Yeah. So like that's something that a lot of people are working on is like, how do we make better semiconductors using ai?

How do we make [00:33:00] things more efficient using ai, AI for design? Within that space we're like, yeah, AI will help. And then it's a, you know, self punctuating cycle. Um, I think that's gonna be. A big piece is like, how do we make better, more efficient compute? And that's why Nvidia, I don't know if they're still as of today, the most valuable company in the world, but they have been for the last year or so.

Yeah. The most valuable company in the entire world. And they make the com, they make the computers that process ai. Yeah. And so whether it's that or whether I think data is gonna continue I also think that with models being released, like, like DeepSeek just got released a couple weeks ago and I like it, it, what it looks like is that they're able to have comparable, if not better performance than Chat GPT for a fraction of the compute cost and compute resources needed.

So like what does that look like as, and that as, as that kind of rolls out as like how do models become more efficient? How do AI systems become more efficient so they, they don't need to use as much power? I [00:34:00] think that's gonna be a huge change too. And something that is constantly being pushed 'cause.

Not everything that you might want AI to support has access to the internet. Imagine like if you were putting a sensor or you were putting something at the bottom of the ocean or on the moon or in the middle of the rainforest or in the Arctic, or like just somewhere on your, since you're at an oil refinery that has a site that has limited cell reception, you might wanna have a fall detection if you're on an oil refinery, you might wanna have like a fall detection system where it detects somebody who falls over or gets hurt, but you might not have internet access.

So you need that system to be able to, you either need to then, in that case, set up your own server on site, or you need to have what's called edge computing, which is on device ai. And oftentimes thosey, those models are much smaller than you might build if you're using a big one. Um, or you might have access to, to, you know, bigger compute resources.

[00:35:00] If you can get models smaller and smaller and smaller and more and more efficient, I think it opens up a whole new world of like, where we can have these intelligent systems. Yeah. Um, around the world. It's not just gonna be in your, you know, downtowns of the world. It could be anywhere.

Kelly Kennedy: That's amazing.

And I, you know, I wanna spend a little bit of time on this particular subject with you because all that I've really experienced as of date, like this date really has been large language model ai, right. Like Chat GPT chat bot, we'll call it chat bot.

Adam Danyleyko: Yeah.

Kelly Kennedy: How, like, what else is there and how long until I can have just like a real conversation That sounds like another human chatting back to me.

Adam Danyleyko: Oh yeah. I feel like you probably have a real conversation with someone to today. Like Chat GPT has the, the voice ch have a con, like a voice conversation with Chat GPT. So like, I don't know how, I haven't tried it, so I don't know exactly how human sounding it is, but. There's lots of uh, there's lots of ai text to speech, to text, [00:36:00] uh, type systems, deep fake audio, deep fakes, things like that that are available today.

Now, okay, so what was the first part of your question? I answered then first part.

Kelly Kennedy: You know, I mean, I'm only familiar with large language language models and I think as a, as are most people as of this time.

Adam Danyleyko: Yeah.

Kelly Kennedy: Is there more to it than that? Is that just one use case?

Adam Danyleyko: Oh, yeah. Large language models are one tiny fraction of, of AI machine learning.

So, I mean, like things you already talked about, like your spam filter on your Gmail account, that's been AI for a long time. Like how, like, like, you know, how, what are some of the things you're interacting, like your Amazon, you recommend your systems like for Amazon or, or Netflix, but like. Everything. Like, you know, apple just released Apple Intelligence on your phone.

So now Siri is power, like Siri overall is an AI system. Your at your Amazon home, your Alexa, all that kind of stuff. And there, those are all a magnitude of different AI systems all working in collaboration with each [00:37:00] other. Um, but there's, yeah, there there's honestly like anytime a prediction is being made, an AI system can do that of some kind.

So like some of the projects that we've worked on here that are, that are public and we can talk about is like, how do we make warehouses more efficient? How do we make delivery routes more efficient? How do we use, how do they use autonomous robots in a warehouse in order to know what order to pick items in to make the, the routing most efficient?

Yeah, it's, there's a lot of applications in supply chain and the financial sector. Fraud detection is a big one, and it's something that credit card companies are doing for a long time. But like, how do you like predicting how socks are gonna do what, what should your investment portfolio look like?

There's lots there. There's honestly, it's like, it's an endless amount of ways that AI can be applied. There's like, again, anytime, like any, think about anytime that you might make a guess [00:38:00] or a prediction about anything, that's a place where AI can help. Yeah.

Kelly Kennedy: Wow. Yeah. Basically what you're kind of suggesting here is that in if not today, in not very long, almost every aspect of our life will have some level of AI integration.

Adam Danyleyko: Or at least could.

Yeah. Like, I think that, that's definitely possible. Like, I think it's, and I don't necessarily think that, like, I think that could potentially sound scary when you first hear that, but it's, it's. Again, like we talked about earlier, is what, how, like, how we look at it is how can we use this as a tool to make your life better, make your, your more efficient, to level up your ability to do work.

And so it's like whether like it's weather prediction or it's root optimization on how you get to work, uh, using Google Maps or how, yeah, how to plan your day better, how to write an email quicker or I don't know. We, we use Gemini Gemini here. And even just being able to like stream of consciousness at [00:39:00] email, uh, and just like bullet point it out and then it'll generate a nice email for you based off of your like, stream of consciousness thoughts.

That's really helpful. 'cause that saves you, I don't wanna say me countless times throughout the day. 'cause I don't have to like first stream of conscious, then edit, then refine. It's just like, I just stream of conscious and then I just review. It's, it's output. Like, it, it, those types of things are, are making me more efficient.

I think that's a lot of ways where when we see, like when I see startups, those are the types of applications where startups I think are really looking at it, is like, how do we take our workers and level them up? Yeah. And that huge amount of the startups that we're we're seeing is seeing applications for, is they're, they started small.

They maybe were like, let's, for example, like they're in Edmonton. Complete random examples. Like they're in Edmonton, they're doing supply, they're IV deliveries in Edmonton.

Kelly Kennedy: Yeah.

Adam Danyleyko: But when you're just doing Edmonton and you have 10 drivers, Kelly can all coordinate the 10 drivers and know which routes to do each day.

Yeah. But when you go from Edmonton [00:40:00] to all of Canada. How many Kellys are we gonna need to hire in order to do that, versus you can build an AI system to help optimize those routes. And so you're able to e like you're able to expand and hit that growth Yeah, a lot faster, um, than you would if you had to physically staff up that many different positions.

So it's not necessarily reducing the amount of staff that startup needs, or they, they're not reducing their headcount, but they can grow faster than they would have to grow their headcount. Going forward.

Kelly Kennedy: It's a force multiplier at the end of the day. So it takes one for sure. Brings them into three.

Adam Danyleyko: Yeah. Right.

Kelly Kennedy: Yeah,

Adam Danyleyko: totally. Yeah.

Kelly Kennedy: Yeah, that's pretty incredible. Pretty incredible.

Adam Danyleyko: And like, and you talked about how you're leveraging it within your own business too, to multiply yourself because like, like you said, like if you had to do all these different pieces, you have to have a human doing all different pieces you're doing, you might need a team of two or three other people to, to support you.

Kelly Kennedy: Easily. Easily.

Adam Danyleyko: Yeah.

Kelly Kennedy: And you know, and that's growing all the time with the scope of change. But like the things that you can use AI for are also growing all the time. And so [00:41:00] it's, it's really interesting. And that was kind of one of the questions that I had for you. And I know it's funny 'cause this show's coming out about a year later after recording it, but.

Talk to me a little bit about what you guys are seeing at AMII, like you are on the cutting edge of ai. We only see it when it becomes, you know, commercially available. But what are some of the cool things that, uh, you're seeing that you're allowed to talk about? I know that you're not allowed to talk about all of it.

Some of it is, uh, is under wraps, but what you are allowed to talk about, you know, what's coming, what do we have coming?

Adam Danyleyko: Yeah, so I would say like I, I'm not not allowed to talk about things be, uh, for anything other than we have NDAs with clients, but, uh, I can talk on, on a conceptual level. So I think something that we're seeing a lot of right now is, um, reinforcement learning as applied to process controls.

And so how can we use. That tool, the tool of reinforcement learning, which as a reminder is like an agent is in a dynamic environment, makes a decision, gets feedback and adjusts how it makes decisions moving forward. How can we use that in process controls for whether it's [00:42:00] from you know, anything like manufacturing to pro uh, oil production, to water treatment, those types of things.

Uh, how can we leverage that tool that that's a really unique opportunity and something that we have a unique expertise in here at a e. But another thing that we're seeing more of is, we see a lot of right now is how do, how can company companies leveraging large language models in a multitude of different ways.

AI agents and ai, LLM agents are, are a very popular thing today. We'll see if there's some popular in the year. Yeah. Uh, but that's something that, that we're seeing a lot of and a lot of interest in is kind of age agentic applications of large language models to do a series of tasks or more like, yeah, like a, a, a workflow of tasks, uh, all at once without human insight and human insights.

So it's like you can give the system, uh, a task that's gonna do a bunch of different things and then come back to you. Those are things that are really popular, kind we're seeing kind of building right now. But large language models are a huge part of what we do and, and how we, we, you know, [00:43:00] different companies are able to, to leverage it.

I think, you know, it's a factor of it being in the zeitgeist right now. That's. With Chat GPT being so popular and so all everywhere people are seeing that application of it. But there's a lot of other ways that companies are, are leveraging ai and it's often in ways that are very unsexy.

It's ways that are very behind the scenes that like your customer would never know of. It's just like it's helping the business run more efficiently. It's helping things run smoother or faster or, or, you know, more efficiently. Yeah, it's, there's a lot of really cool applications. We, we are working with a ton of different clients right now in terms on the buildout side, but on the the startup side, we work with hundreds of startups a year, helping them see where AI can be applied to their business.

A lot of those startups are coming to us. Not knowing where to start. Yeah. They're like, they don't know where AI can be applied to their business. They don't know what, how it can look or what the different op options are. And we're helping them identify those opportunities [00:44:00] within their own business. And what I like to, when we're talking to them, what I really like to start with is start with your business problem.

So like a lot of our workshop for that is by helping them think about where are the opportunities in their business where they're seeing a pain point. Like is there like a roadblock, is there a pinch point in their processes? Is there like a human, like a, a capital constraint on the human capital side where it's like, this is a, a pinch point for growth or for where we're currently at?

Is there a slow process or something like those, those opportunities from a business side are like, those are problems that we need to solve.

Kelly Kennedy: Yeah.

Adam Danyleyko: Now is there an AI application that can help us solve that problem? So I think that's the better approach. Like, what's the problem? Let's find the right tool to solve it.

Rather than, you know, we, we still see a lot of companies coming to us saying, Hey, how do I use Chat GPT in my business? Yeah. Because like, yeah, I mean, like that's, maybe a good question, maybe not. 'cause it's like you wouldn't walk up to a construction site and say like, Hey, I have a, a new screwdriver.

What can I screw in with this? Right. [00:45:00] Like you, it's the same thing, right? It's like I have this, there's this shiny new tool available. How do I use, how, how can I use it? Rather than, here's the problem that I have, let's find the right tool for the problem.

Kelly Kennedy: It's a really interesting conundrum when you split it up between startups and existing companies because, because I think on a certain level as a startup, being able to start a company and incorporate AI from the beginning is a massive, massive advantage over having to figure out how to integrate AI into your already existing business.

You know, let's talk.

Adam Danyleyko: Yeah. I'd say yes and no though, to be honest with you. Like. Yes it's really cool because you can build it into the DNA of your business right. When you start.

Kelly Kennedy: Yeah.

Adam Danyleyko: But it's also sometimes really challenging because AI needs lots of data.

Kelly Kennedy: Mm.

Adam Danyleyko: And so if you're a new company, you probably don't have much data to start with.

Right. So, like, oftentimes, like we're, we're, we, like we're working with companies that, that even so on the startup side, like they, they have a product in market. They've been, they've been collecting data with their cus from their customers. So I think it's [00:46:00] important to, like if I was giving advice to someone who's starting a, a company right now, it would be think about AI and think about how you're going to leverage ai.

And if AI is gonna be a fundamental piece of your business, then you need to start thinking about a very, like right at the start. But if it's gonna be something that enables you to grow and enables you, like down the road, start collecting the data you need to power that AI system now. Yeah. Even if the system isn't gonna get built for a year or two or five, but if you don't have the data and in five years.

Now you're starting, like now you're, you're five years behind the ball. Yeah. You could be collecting the data that you need now for that eventual system in five years. But there's all, I mean, not if you're leveraging large language models, you don't need as much of your own data. 'cause they ha they trained it.

I guess it depends on like. For the, like a large language model application, you might need a knowledge base. Yeah. Um, it might be like, you know, a theoretical knowledge base or something else that, that you're helping to additionally train it on as [00:47:00] opposed to needing sensor readings over the course of five years.

Kelly Kennedy: Yeah.

Adam Danyleyko: Which you might need for different applications.

Kelly Kennedy: Absolutely. Absolutely. I, I get what you're saying. I think, at what stage should a company start thinking about incorporating something like a chat bot or a large language model into their systems? Is it something that companies should be doing from the very beginning?

Is there a stage, like what is the cost of incorporating maybe a system, let's call it, to a small to medium sized company? Is it, is it like drastic? What does this look like?

Adam Danyleyko: I think it, it honestly, like the number one answer I think when it comes to AI is it depends. And so that's something that we say all the time is like, it, it really does depend.

Which is like, maybe not the best answer, but I'd say it's probably the most honest answer I can give. Sure. Yeah. So like, there's definitely, like, you could probably, you could potentially turn around, uh, a simple chat bot, for fairly inexpensive using A-A-G-P-T or similar API. You can also, like, if you want something though, that is, and that's like for, you know, [00:48:00] basic question answer type things.

But if you want something that's trained on, like, let's say on, on your podcast and you wanna say like, Hey, can we even build a chat bot that's trained on my podcast and I can ask it questions about, you know, yeah. You know how, you know, like. Who's talked about ai, uh, on the podcast, and then it's like, what are the different topics we've talked about?

How long have we talked about this? Like, oh, you know, you could really but true on your podcast. That would take collecting all that data. Yeah. And then training it probably using a, a what we would call a rag system or retrieval augmented generation system that is paired with a large language model in order to work together on that.

And that that application is gonna be like, more complex to build and more costly. And then think about like, instead of maybe one podcast, just, you know, just this podcast. But what if you're training it on government Alberta's data? Yeah. Or like Proctor and Gamble's data or something. Yeah.

It's gonna be like exponentially more expensive and exponentially harder and more difficult. To do so it, [00:49:00] the scale depends the amount of things you want it to know and to be able to answer is gonna influence the ultimate price, um, cost, uh, to implement it. The duration, how long, how many people are gonna be required to build it.

They can, like, it can be honestly from very cheap to almost nothing to as expensive as you wanna, as you possibly wanna spend. Yeah.

Kelly Kennedy: How nice of a Ferrari do you want?

Adam Danyleyko: Exactly.

Kelly Kennedy: Yeah. Yeah. No, I get it. That actually makes a lot of sense because let's get real, even developing a website can be fairly cheap to incredibly expensive depending on what it's you're building and how big it's, that makes perfect sense.

Adam Danyleyko: And with AI it's a lot more different. It's a lot different than like building a website. Like building a website is pretty, pretty standard. Yeah. You, there's different ways to do it. Yeah. But it's like there's a way a pretty understand should wait to do it. Just like, how much time does it take, how much energy does it take to do it?

Effort With ai, a lot of times it's like we're doing experiments. Yeah. Like we're trying things and we're seeing what works and what doesn't work. And whenever you're experimenting, there's [00:50:00] no guarantee of result. We like, like when we're doing stuff with our clients, it's a proof of concept development.

And so like that often means like. Like I said, we're doing experience. We don't know what the right approach is right off the start. We don't know exactly what the results are gonna be. We have an idea of what it should be.

Kelly Kennedy: Yeah.

Adam Danyleyko: And we have targets that the client is aiming for, but it's, we need to figure out what the best approach is.

We need to figure out which is the right model to use. We need to know what, like what data should or shouldn't be used, how we should process that data, what model is needed to do that, that calculation. All of that is an experiment that we're doing to figure it out the right or best path.

Kelly Kennedy: Yeah, absolutely. I know we have a lot of companies listening right now who might be asking themselves, well, how do I know, how do I know that AI would support me with my challenge in my business? Right. Because when you're coming to it and that's not your background, it's really hard to know whether or not an AI solution would be the right solution.

Yeah. Um, can you maybe just gimme some examples of success stories from [00:51:00] some of the companies you've worked with at AMII and maybe that'll give people some ideas on what they could do with it.

Adam Danyleyko: Yeah. Um, I think that's a really, really good punch. I would say AI is not the solution for everything. Like there, there's a lot of solutions where it's like, yeah, if you can do it a good solution without ai, like don't build an AI system just because you can like it.

They, like you said, they can be expensive, they can take lots of time, and they require maintenance and upkeep as they're going. So when they work, they're really good, but they're not the solution for everything. It's kind of maybe my first kind kind of ca out there. Okay. Um, this warehousing one that we did, we did some, some generative warehouse design in order to make warehousing more, more efficient and and bad.

Like, so you, that's a, a combination of, of generative AI in terms of design. That's been really cool to see how that, that kind of rolls out into the real world. We've seen applications in, yeah, like, that'd be like a supply chain application, but in medical devices in terms of detecting different conditions [00:52:00] or providing insight on like, you know, different levels.

Like readings, try, I'm trying to be as specific as I can without, uh, going into like, you know, super duper detail. It's a really cool example here, here in Edmonton, um, is a company that a company called, now they're called Meadow ai. So they were orig, no, they were Meadow ai. They founded in Edmonton.

And now they're worried they got, they kind of joined with a US company called Echo. So now they're, they're called Echo, but they do machine learning on ultrasounds. And they like on like ultrasounds that can connect to your phone or an iPad for things like, um, they start with infant hip dysplasia.

So they're able to scan that infant's hip. To determine if it, if they're at a, a high risk or a low risk for hip dysplasia. Hip dysplasia is something that if it's caught in, infants can be corrected while they're an infant and cause and if, if, and it stops them the, when they're an [00:53:00] adult from potentially having severe arthritis in their hips and potentially needing early hip replacement.

So this is something that like, has been screened in large urban centers for a long time because they have the devices and the capability to it. But when you're able to then pair that machine learning tool with a, a device that is able to be, again, attached to a phone, then you can take it to rural communities a lot easier and scan way more people around the world.

Yeah. Uh, and stop that condition in a way that has a real impact on, on the world and on people in rural and indigenous communities, um, you know, across Canada in a way that is. You know, it's having a real impact on people's lives. Yeah. Um, and it, it allows also, like, so my, my, both my kids have been scammed with that device at our doctor's office.

Wow. You, it was so cool the first time I went to get my daughter and I was like, I know those people, like, I know made too. Um, it was, it was so cool seeing it and it's so easy, like the nurse who maybe doesn't even ha like normally wouldn't have the [00:54:00] training to to do this test, you might have to go to a specialist to do it.

Then now the nurse at our local, our doctor's office is able to do this test. Um, and so like things like that are, are really cool applications. Wow. Um, that, that we've seen. But like I said, with a lot of the startups, the best application, I think the most common ones are things around how do we take this process that is working manually up until our point of rough scale.

That is gonna start breaking down. So what are those processes that start breaking down as you're scaling that an AI system can be used for instead? A set of a manual process. That, that's kind of where I've seen a lot of the most success with our, with our startup community. But there's so many different ways that can be applied, whether it's scheduling or routing or validating versus like validating invoices or building invoices or, or, there's so many different ways that, that those, there's so many manual processes in a business.

Yeah. That when you're 10 people looking at this geographical region, they, you know, you just do it. Yeah. [00:55:00] When you're growing to be like 500 people across, uh, a country, it just, those processes break down.

Kelly Kennedy: Yeah.

Adam Danyleyko: I think that's where, that's where the, I've seen startups have the most success.

Kelly Kennedy: Yeah.

That actually makes a lot of sense because especially if you have to train up an entirely nother, like an entirely other group of people in the next city. When potentially you could just train up an AI-based system and teach everybody how to operate that, you're gonna have a lot less errors, a lot less challenges, and things are gonna flow smoother.

It makes a ton of sense. Okay. If people are listening right now and they need support, this is something that AMII can help with and not just in Canada. Can we chat about that?

Adam Danyleyko: Yeah, so we work with companies all around the world. And so no matter where you are right now, listening, like if you wanna learn more about how we can help you reach out AMII.ca is our website.

We're based in Edmonton, Alberta here in Canada, but like I said, we work with companies all around the world. We happy to chat with you. There's a, we have an intake form on our website, but if you fill that out, we'll someone from our team will get in contact with you, um, and set up [00:56:00] a, you know, get some more information from uc if we have a program to support you how we can best support you.

Uh, on the startup side, we have free programming for Canadian startups. So that is either through our machine learning exploration program, which is to help sort of see where AI can be applied to their business. Or through our Level Up program, which is for AI startups, you know, startups that are already using AI in their product, provide them ongoing coaching and support as they scale.

Again. Now those are free programs for Canadian startups application based, but free if you get, if you get in. Yeah. Um, but all of our other programming is a fee for service model and we're, we're happy to work with companies all around the world.

Kelly Kennedy: Amazing. And Adam, if people wanna get ahold of you, do they just go to the a e website directly, or should they follow you on LinkedIn?

What's, what's the best way to do so?

Adam Danyleyko: Yeah. Three, follow me on LinkedIn. Going to the AMII website is probably the, the best, uh, the best and easiest way to get ahold of me though, if, if you're looking to, uh, to get ahold of me personally.

Kelly Kennedy: Amazing. Dude, this has been incredible. Thank you so much for [00:57:00] helping us with the Alberta Ecosystem Series.

Like I said, you know, we're incredibly fortunate to live in Alberta. We probably have some of the best entrepreneurial supports in the world, but sometimes they're really hard to understand. So having this conversation, letting people know what you can do, how you can help 'em, what you know AMII is massively beneficial.

And thank you so much for your time today.

Adam Danyleyko: I know, I appreciate having the opportunity to come back here again and have, uh, I, I don't remember exactly what we talked about last time, but I feel like it was a similar but maybe slightly different conversation.

Kelly Kennedy: Always. But I'll tell you what, in my experience, round two is almost always better and I can pretty guarantee you we had a this time anyways.

Adam Danyleyko: Well, you get the, uh, kind of the, the first conversation jitters out. We, we know each other a little bit better now. Right. So. Yeah.

Kelly Kennedy: Exactly. Exactly.

Adam Danyleyko: No, it's good. I appreciate it too. And, and it's always, uh, it's been a pleasure chatting with you both times and I'm, I'm excited to, to, I'm always excited to talk about AI and talk about AMII.

It's a great, uh, great place to work and there's really cool things happening here and I love working with startups and [00:58:00] working with different companies as well and helping them grow, uh, and helping them be the best they can be.

Kelly Kennedy: Amazing. Well, you continue doing that and I'm sure lots of people hear this and reach out, so thanks for your, uh, thanks for your help, Adam.

Thanks for what you do and we'll chat soon. Yeah, thank you. Until next time you've been listening to the Business Development Podcast, and we'll catch you on the flip side.

Outro: This has been the Business Development Podcast with Kelly Kennedy. Kelly has 15 years in sales and business development experience within the Alberta oil and gas industry, and founded his own business development firm in 2020.

His passion and his specialization is in customer relationship generation and business development. The show is brought to you by Capital Business Development, your Business Development Specialists. For more, we invite you to the website @ www.capitalbd.ca. See you next [00:59:00] time on the Business Development Podcast.