The Future of AI and Data Analytics Explained with Sean Byrnes of Outlier

DISCLAIMER: Below is an AI generated transcript. There could be a few typos but it should be at least 90% accurate. Watch video or listen to the podcast for the full experience!

Sean Byrnes 0:00
Even today with artificial intelligence is all our progression of technology, even talk to your video companies are made out of people, it’s still the case. As long as that’s true, the best thing I think you can have is a really, really clear idea of how you want your employee experience to be and what you want to treat them as.

Alexander Ferguson 0:23
Welcome to UpTech Report, this is our apply tech series UpTech Report is sponsored by TeraLeap. Learn how to leverage the power of video at teraleap.io. Today, I’m excited to be joined by my guest, Sean Byrnes, who’s based in Oakland, California. He’s the CEO at outlier.ai Welcome, Sean, good to have you on.

Sean Byrnes 0:42
Thanks for having me.

Alexander Ferguson 0:43
Now, outlier is an automated business analysis tool. And basically automating what your a business analyst would do if they had the time. You guys are focused on enterprise space, large consumer businesses are your main space of those who have millions or hundreds of millions of customers working in retail, hospitality industry pharma supply chain, often starting with the CMO. So if your cmo there, it may begin with you, but others could use this business analysis tool. Help me understand those, Shawn, what’s the the problem that you you see in the marketplace and really are working to solve?

Sean Byrnes 1:16
Absolutely, these large consumer businesses have so much going on US consumers are changing every day, their buying behaviors and demographics. It’s very hard to stay on top of that, to know what questions to look for to know where the emerging problems and opportunities are hiding. Because your data is distributed across dozens of different systems out there connects all the systems looks through that data and brings you those questions you should be asking those emerging problems and opportunities automatically. So you can do what humans are great at which is make decisions and be proactive.

Alexander Ferguson 1:46
I love always focusing on more what what humans should be doing and what we don’t really want. Let’s let’s go back though, how did outlier begin what was the origin so.

Sean Byrnes 1:58
So I’ve been entrepreneur for a long time, my cup, I’ve started a company before this one called flurry, which is a large analytics platform for mobile applications. So if you’ve used iOS apps or Android apps on your smartphone, there’s definitely flurry somewhere on your phone. We were the largest service provider, we sold it to Yahoo many years ago. But through the growth of mobile flurry was the leading provider of analytics, which is great, it means I got a front row seat to how hundreds of 1000s we had about 500,000 customers to the point in which we we got acquired. So I got to see how fortune 500 companies were using data how individual small businesses were using data, how all these businesses were using data to make decisions. And I was lucky enough to meet as many of them as I could in different countries and different places. And a funny thing started to happen, which was that as I would go to these companies, it doesn’t matter if they were small or big, or in the US or Europe or Asia, or in consumer, whatever they were doing. They all started to ask me exactly the same thing, which was Shawn, I love this data. This is great. But how how do I know what to look for in all of it? How do I find the opportunities? How do I find the problems? And at first, it sounded like a naive question. It’s like it’s data, you look through data, is that what you’re supposed to do. And eventually, I got so often that I realized it was actually really profound. And what they were telling me was that the data had gotten to be so big and so vast, that all of their old ways of thinking about it creating dashboards and spreadsheets, they just were they were breaking down. They weren’t working anymore. And we needed a new kind of approach that would help us find these sorts of questions. And that became the impetus for outlier, which is over the next decade. What does it look like to give us a new set of tools, who instead of like traditional business intelligence, which is great at answering questions, we don’t ask, what does it look like to have tools that bring us questions proactively that don’t wait for us to tell it to do something that they’re looking through our data, wherever it is. And again, it’s going to be in dozens of places, let’s be honest, it’s not in a nice deep pile in one place. And bring us these questions that we can be more proactive. We can identify these customer the shifts in customer behavior early we can identify demographic changes or market conditions to take action and make better decisions that became the impetus for Allah. But I should tell you, when we started the company, I actually didn’t know if this was possible. This is the problem. But I didn’t know if you could actually do this, uh, could you write software? And I’ll tell you as I would explain to people what I wanted to do, as I’ve explained to you, before I had a product, inevitably I get these funny looks. And people would say, first of all, that sounds like science fiction, there’s no way software can analyze my business, my business as a snowflake is not possible. But the other thing that they would say is listen, I just heard a bunch of data scientists, they’re wearing white lab coats, they’re gonna solve this problem for me, I have this nailed. And it’s kind of amazing how over the last five, six years things have changed that it’s just come so far. And now this category is seen and accepted as the large the future where business intelligence is going, but it’s been quite arrived. So

Alexander Ferguson 4:55
you’re definitely not new to business intelligence and data, as you said with flurry For 889 years close to that, we’ll come back to again of how outlier is growing. So but let’s take a second though. How did you get into entrepreneurship? We? Are you just like naturally ready to start new things? No,

Sean Byrnes 5:19
no, actually, to be honest with you, entrepreneurship was a lot the farthest thing from my mind, I would. The thing to know about me is I’ve always loved doing lots of different things. For a while I wanted to be an artist. I wanted to go into academia become a researcher, I love building things. I wanted to become a woodworker, there was no shortage of things I wanted to do. I’ve never been the best in anything. But I’ve always loved doing a lot of different things, and specifically the interplay of how these things relate. And

Alexander Ferguson 5:50
in the back, by the way,

Sean Byrnes 5:51
yes, yes. Yes. The artifacts of my previous life as an illustrator. Yes, yes, most of my illustration today is just done for my children. But yes, a long time ago in a foreign place. But you know, there’s it was always fascinated the interplay between art and engineering, or, you know, music and writing and communications, and these sorts of things. That’s always what fascinated me. And I found that as I was doing things, I kept getting bored. Because everything that I would do, they always wanted me to do one thing and keep doing it, doing it doing it. So in academia, they wanted you to do one kind of research and go really deep, or what I have a job, they will show up for your job, they want you to do your job. And that’s what you do. And I really wanted to be challenged other ways. At the same time, I also started to have this weird feeling that all the companies I worked at, they were they don’t see their employee, they didn’t see their employees as people, they saw their employees essentially as widgets, where you put a salary in one side and out the other income, some for a productivity. And often they would tell you what to do and how to do it. And your job was to go do that. And I wanted to grow, I wanted to make mistakes and learn and expand. And I just felt very limited. And so when I found that entrepreneurship was a chance to create the kind of job that I want it and the kind of job I thought other people would want, which I know is a very strange reason to go into founding companies, or if people typically become a founder, because they want to make money or they have dreams of building the business. I just wanted to create a job that I could love and other people could love as well. And that became the motivating factor for my first company for this company, every company I’ve started that has been the motivating factor is can we create a place that people would love to work? Because my thesis is simple. If you create a place that brilliant people, smart people want to work and they love working at everything else will work out, it will create value to be valuable. It’ll solve these sorts of things. It’s important. Now, that being said, you have to pick a good problem to kind of point those people at which is where this impetus for outlier came from. But yeah, my motivations always been the people. Because I think that there’s so much dehumanization, that happens in modern companies, and so few businesses are willing to treat their employees, like people are like adults, that it shouldn’t sound like such a radical idea to start businesses that really are just aiming to treat people like adults. But it turns out that it is

Alexander Ferguson 8:11
this interesting combination of of I mean, you started, at least I see on LinkedIn, a software architect working Verizon, but then you have this design and combining engineering with design fascinating combo, but then you jump as you say, start your own business, because you want that freedom to be in your own job role that you feel fulfilled in others. But have you when you started your company? Had you managed a large team? And was that a new learning for you?

Sean Byrnes 8:39
Oh, no, I definitely not. But I can tell you, there’s little bits and pieces that you put together along the way. It’s when I started my first company, I’d never done anything like it before. But you know what I had done. I had I had been a DJ on the radio, I loved music. I did that for a while. So I got really great at public speaking. So it turns out, I was great at explaining what we did. I went to graduate school for artificial intelligence and computer science. And so I was great at building software. I had spent a large part of my life in art. So I understood composition and how people understand images and understand context, those sorts of things. And I had along the way, also work, I’d run some large volunteer organizations. And so I hadn’t ever done that. But the PA turns all these disparate pieces, they don’t sound related. They came together, and we’re a formula that actually helped me thrive. They helped me do a lot of things that I think other people might struggle with. Because while I was able to do a lot of these little things to help the company get started, I never felt like I was the best at this. And so I didn’t want to hire somebody to do it for us. So as the company kind of got going, I had no problem giving up the different roles to people because I was like, I’m not the best at sales. I can sell. I can talk I got very good at that. But I’m not the best and so hiring somebody to do that is okay. Same thing with the engineering. I can build the software but let’s be on So I’m not going to be the best that that as soon as we get to a certain point, let’s bring in people who are the best at that. And so that helps as well. But overall, I would say the most important thing was that that Northstar of starting a company where the people are valued as people first was the most important decision because it defines how you grow the organization, the kinds of people that join you how you treat them. And let’s be honest, even today, with artificial intelligence at all our progression of technology, even talking to your video companies are made out of people, it’s still the case. And as long as that’s true, the best thing I think you can have is a really, really clear idea of how you want your employee experience to be and what you want to treat them as. So all those pieces came together. But I still think that was the most important observation that I had from the beginning.

Alexander Ferguson 10:46
It can you paint a picture that you you’re unhappy with, just being working for another company that doesn’t value people, so you decided to break out on your own. You do that with flurry, it grows, it’s amazing. You do it again without like, there’s no problems. Everything’s perfect. You have no issues.

Sean Byrnes 11:06
Yeah, that’s not how it works. It’s, it’s amazing to think every time I’ve started a company is the first one was flurry with outlier. I really thought I knew what I was doing. And I was just horribly, horribly wrong. And to be fair, I think to start a company, you have to be a naive optimist, you have to believe that you know what you’re doing. Because if you understood how catastrophically wrong, you were probably nobody would do it in the first place. And there are lots of people that claim they have a doubt they have a formula and you just look at the success rates of even repeat founders. It’s just not very high. And so you know, Fleury, my first company ended up being a very large success, but it was insolvent twice. Over the life of the company, I got paid for I was paid a salary for maybe five or six of the diet years, it was in business, I it was not, there’s no such thing as a straight line to success. You have to deal with a lot of struggles. And frankly, that’s often the hardest part. And in some ways, my motivation of building a business for people was one of the things that helped me get through those dark days. Because if your goal is starting companies to become rich and famous, as soon as you hit that first very difficult spot where everything looks bleak, and it looks that there’s no path to success. And you by the way, as a founder, you go there pretty often, you get very comfortable in that place. You’re going to give up because it doesn’t look like you’re gonna get rich and famous from this company. So why would you keep investing in it if that was your motivation. But the good news is that if you have another kind of motivation, like people first or whatever it might be, you get you get through those, you kind of can see a light because you can see your purpose and you can keep putting one foot for the other and get through them. But yeah, building companies is not linear. It’s not. It is a roller coaster. In fact, I would say that the one number one cause of failure most early stage companies is founders being unprepared for the emotional rollercoaster, the highest or intoxicating, the lows are so difficult to overcome. And they come in such rapid succession that you’re yo yoing back and forth between being ecstatic and being depressed and being sad and being depressed. And it’s tough. It’s actually in some ways a learned skill to deal with the the amazing emotional range that you go through on a regular basis.

Alexander Ferguson 13:17
Is there anything that you found works well, for you to ground yourself in that amazing up and down flux?

Sean Byrnes 13:24
It’s tough to say I will tell you it flurry, I don’t think I did it very well, I had a lot of very unhealthy habits. I’ll tell you one thing I’ve dealt with outlier that’s been much better is that I’ve realized your mental health is a big factor in your productivity and your success. And so I do things now that I didn’t use to I’ll give you some examples. I used to feel guilty if I took time off, because I’m like I should be working I should be things forward. And now I realize that taking time off and sleeping well at night. That’s part of my job. Because of I am better rested, I can make better decisions. I used to feel guilty when I would go work out. But now I’m like exercise as part of my job because the better my fitness the better I can deal with these stressors. And so I’ve started to take things that I used to feel guilty about which were very negative, destructive patterns. And I think about them now as part of my job, where my job is to be well rested in shape. And so I can make these important decisions when they come up and deal with these moments because I know that they’re going to happen. Coming into my first company at flurry, I don’t think I really understood what it would be like and so I didn’t know anything more. I mean, then what when you think about it, what do you do in grad school, you’re just hustling you’re pushing to try to get through things. And that was the coping mechanism I had. And it’s that one doesn’t work very well, because eventually you can’t just push through these things over many, many years. So I did that differently. It’s worked out quite well. I will say this that I’m able to enjoy the journey a lot more. I don’t feel like I’ve lost anything in terms of the ambition or the energy or the productivity. But I do feel it’s much more sustainable. Like I think people realize that it’s not sustainable and there’s a lot of people out there that brag about it. sleep for hours at night or I work seven days a week. And I actually at one point, I did have a friend, who, who else is a very interesting conversation told me that if he hadn’t seen his kids all week, his wife would bring his kids to work on Sundays. So he can see them at least once during the week. And that’s an extreme example. But you know, you think about it. That is that is a place a lot of people exist. And I understand why they do it. And I understand why it happens. But I fail to believe that as an this that is necessary to kind of be as productive as I think we all want to be.

Alexander Ferguson 15:38
You for both companies, are you a single founder, co founders.

Sean Byrnes 15:42
I’ve always started with co founders. I’m the kind of person that works better in team. I think I need people to help me refine my ideas to keep me accountable. I live in all solo founders to be honest with you, because the amount of internal drive and focus you have to have is immense and impressive. And I I prefer to go with other people. I’ve always been that way. I prefer team sports. I prefer you know, working in groups, and I prefer working with co founders.

Alexander Ferguson 16:13
How did you find your first co founder? And were they different people both in companies and what did that look like?

Sean Byrnes 16:19
They were different people. My co founders were flurry were people in college, as it often is in those early days after graduating. So that was on surprising. My co founder outlier, I decided to take a different tact, which is I realized that in a lot of ways, the best co founders are not going to be your best friends. They’re gonna be people who have similar working styles, they have similar skill sets. But they also the most important factor internalizes they deal with stress the same way or more specifically, very predictable reaction to extreme stress. Because again, you’re going on this roller coaster together, how do you deal with that extreme stress if the person you’re working with you don’t you can’t predict how they will react. And by the way, this is extreme stress that can have a lot of money on the line as well, which brings out all sorts of surprising behaviors and people. You don’t want to be surprised. And so I had a process of looking for co founders and talking to them, interviewing them, and that’s what rigorous processes really stress testing what it’d be like to work together. And I’ll just say my current co founder Mike Kemet, outlier. He had never started a company before. But he’s done amazing. He’s done a fantastic job as a first time founder. He’s dealt with everything really well. He’s excelled in areas, he’s made the company more successful. He’s definitely the core of what’s made us most work. And I think a lot of it was the fact that we do work well together. We have similar working styles, and we do deal with stress in predictable and understandable ways. And so that has been a bedrock of the business. And you know, I said before that the largest cause of death is the dealing with this, this kind of emotional rollercoaster that manifests itself and in founder disputes and founder fights and founder Fallout and founders leaving, and we’ve, we’ve have a strong relationship today as I think by stronger than when we got started. So that’s worked out very well. But it was a we spent a good I don’t know, a few months, trying testing the waters testing working together before we were confident it could, it could work.

Alexander Ferguson 18:23
You listed yourself as a CTO and flurry. Are you a tech person? You’d love technology? Like is it something you enjoy?

Sean Byrnes 18:33
Yeah, well, I had almost every job but flirty, I was a CEO. And then I was a CTO and I was Chief Product Officer. I had lots of jobs. I mean, if you’re a founder of a company, it lasts that long. They’ll eventually do everything. It’s just how things work. I do love technology. I started I remember certain deliver program in high school or my parents PC. And when I went to college, one of the first classes that I took was computer science got really fascinated about computers. I didn’t love it though, I really much more enjoyed I actually the major engineering because I much more enjoy building things that I can hold and and see the interesting turning point for me and technology was I took a class my junior year of college called machine learning or AI was called artificial intelligence. We’ve gone back and forth about what we call it. And and I remember being in this class and being amazed by the idea that you can teach a computer to do things like that just that concept was just mind boggling to me. At the very end of that class. And the last day I walked up to the professor and said I want I want to do research. This is what I want to do. And I wasn’t an edge. I was not a computer science major. And so her name was Dr. Daniela ruse. She’s fantastic. She took me on. She let me do research in her lab for my senior year. And then I was like, I want to go to grad school and learn more about this. And so I went to grad school for for machine learning artificial intelligence, and just remained fascinating even to this day. It’s amazing to me that we can build things that we don’t design that we teach And they’re very primitive. Like, you know, we can’t they’re not like children people talk about general intelligence and let’s be honest, that’s gonna set so far off. We can barely keep things laid and if I images, but the fact that we can even do what we do is just mind boggling. Still, to me, I have so much offer this kind of technology, this kind of space what it can do, because it feels like magic. It feels like the closest thing to magic that I think that we’ve made in recent past because you know, even a product like outlier, people use it. It feels magical. It’s technology. But I imagine you imagine what what was it like the first time someone saw an airplane, and you’ve never heard of an aeroplane before, but you saw this object take off and fly. And nothing in your life experience have prepared you for a machine that can do that. That kind of magic. I feel like we’re seeing today with machine learning. And it’s just it’s still exciting to me today.

Alexander Ferguson 20:51
What year did you go to grad school and machine learning.

Sean Byrnes 20:58
So you should know that I have made enormous li bad decisions throughout my life. And this was one of them. So I was in college, I graduated from college in the year 2000, which was the peak of the.com bubble. So instead of going out and joined the.com bubble and looking to get rich off of the.com bubble, I was like I’m gonna go to grad school where they don’t pay you anything. This is a great idea. And I ended up graduating as a result in the depths of the rubble. the.com Bubble having burst, right? So I believe in grad school here the.com bubble had exploded. And it was one of the most difficult times in history to find Java technology. By the way, if you remember back then, but it was around. And I was like, wow, that’s fantastic. Thankfully, I did end up letting a Verizon who had a group called E business, which was an incubator for new business models, so they were looking for people that can help them diversify the business of Verizon, which is the phone company, let’s be perfectly honest. But it was in some ways perfect because it was a testing ground where they would come and say, here’s a new technology. So for example, when Wi Fi first became big, how does Verizon make money off Wi Fi. And we got to our team got to essentially build something from the ground up inside Verizon, which was a safe ecosystem. Unlike this founder situation, I wasn’t worried about going out of business or not getting paid, and build businesses that we had the thing that fries and we built this great network in Manhattan where there used to be a payphone in every corner. And we ran, we ran internet to each payphone and put a Wi Fi access point at the top of the payphones. And way back in 2003, New York City had the first city wide what mesh Wi Fi network in anywhere, anywhere in Manhattan, you when you can get Wi Fi everywhere. Oh, yeah, yeah, that was the foundational part of of what we were doing. It was a lot of fun, I got to see how these things were built. But yeah, and going back to what I said before, that was great training for being a founder to like all the little bits and pieces along the way, kind of added up. And that was a big part of it for me as well. But it is funny to think about until outlier many years later, I didn’t use any of that artificial intelligence. I said it in grad school, not even a little bit. In fact, we used to joke in grad school that like what we’re working on would never be used in our lifetimes. Because back then the computers just were so slow compared to what we needed. It’ll be like our children’s children before everything, but then it’ll be amazing, then it’ll all work, then this theory will become reality. And it came a lot sooner than we thought as you can tell. But I somehow I sometimes I think of outlier is like the culmination of the journey. I started way back then.

Alexander Ferguson 23:31
Yeah, 10 1015 years instead of many, many more. Exactly. So you come down to outlier, and it’s your chance to apply it? Did you immediately see that as an opportunity saying, Alright, I’m seeing all these, these connected pieces. And if we could just harness people need insights, let’s use machine learning here.

Sean Byrnes 23:51
No, no, like I said before, we saw the problem that people needed help finding these questions. But how to get there was very unclear. In fact, like I said, before, we weren’t sure we could build anything, forget about machine learning. I didn’t know we could build anything that would do this. And so literally the first say six to nine months of outlier, my co founder and I, we essentially just rented ourselves out as consultants, we went to these businesses saying, Hey, give us all your data. And by the way to get it was in dozens of places, don’t but don’t tell us what you look for. Don’t tell us what your dashboards are, what your metrics are. We’ll go poke around, we’re pulling up and we’ll share with you what we find would be unexpected insights. And let’s see is there is there economic value added, like we suspected is there is going to be something these companies can use to make decisions. And it turned out there was the results of that were resoundingly positive. But then we were left with Okay, cool. Can we build a product that does what we were doing as human experts, and that was still very unclear. And we just also what we lacked was any sort of metaphor, right? If you want to build a CRM system today, you can look at Salesforce and say, I’m going to build a better version of Salesforce because people know what a CRM looks like. I’m gonna build that The version that we were like, What is a quiet? What does a software that ask questions look like I don’t know, like is like Google without a search box. And if you didn’t have a search box or Google, what is Google like, it’s very hard to think about. And so we started building, we started building attempts. And the first few were really embarrassingly bad to be perfectly honest with you, like, really, really bad. But eventually, we started to find models that worked better. And we kept kept going. And I think our fourth or fifth attempt became the great, great, great ancestor, what we do today, but it was a lot of trial and error, it was a lot of imagination, it was a lot of testing, it was a lot of it was a lot, but it was fun. And I think that, you know, going back to my experience, like, you know, being an artist, the most exciting thing you can do is start with a blank piece of paper or a blank canvas, because it’s full of potential and you’re not intimidated by that. And you get very used to starting with no limits. And that ends up being the hard part with most new categories of software is a lot of people struggle without limits, because most of life gives you limits must have life gives you constraints and your job to do the best job inside this box that you can do. But what happens in the box goes away, you need a whole new set of tools for dealing with that I do think there’s an aspect of being more comfortable than venuti. And being comfortable without limits without boundaries that I that I learned very early on that’s helped me in my career.

Alexander Ferguson 26:23
So Sean helped me understand what are your thoughts? How do you get someone to be ready to adopt new technology?

Sean Byrnes 26:29
If adopting new technology means you have to describe or explain or convince somebody they have a problem, then you’re not going to succeed? It isn’t entirely too difficult. But if there is a problem that somebody knows that they have, and when you ask them, if they have this problem, they start shaking you up and down, saying yes, can you help me, please do anything you can to help me, then your new technology can find adoption, because people really don’t care about technology, they don’t really care about methods, they care about solutions to problems they face. And so many people in new technologies, they spend so much time falling in love with the technology, where they really should be falling in love with the problem. Because we fall in love with the problem, technology can change. But the invariant always has to be the problem. Because if somebody knows they have it, they will desperately want you to solve it for them. But if you’re trying to convince somebody, you definitely need this technology, because we’re probably didn’t know you had good luck. Because it’s very hard to convince somebody that this problem they didn’t know they had is more important than these dozen problems that they already knew they had

Alexander Ferguson 27:29
thought here connected to this piece. AI machine learning. It can solve problems. But some people are still concerned are like how can I trust it? How can I know the answers they’re bringing out? Is it real or trustworthy? What do you say to that?

Sean Byrnes 27:42
I think they’re right. I think that AI has a lot of potential for evil. I think that it can perpetuate systemic bias discrimination. I think it can lie and seem credible because it’s magical technology. And I think people right to be skeptical. I think that one of the big mistakes of most AI products today is that they’re black boxes, they expect the user to trust that they are doing what they say they’re going to do. One of the earliest things we needed outlier was realized that you trust is a ladder trust is something you build over time you work your way up, you don’t start saying trust me, this definitely works you earn trust. And so every insight that outlier creates explains to you how it came to that conclusion. It walks you through how it got there. So you don’t have to blindly trust it, you can go check for yourself. Do you agree? Do you think that that was the right way to get there. And by being transparent and explainable, you can earn that trust and get people over that hump. It’s kind of like similar today, self driving cars, right? Most people, if you’re going to a self driving car exhibit, they will not get into there’s not somebody sitting the driver’s seat. And it’s a self driving car. You don’t need somebody the driver’s seat, but their whole life. They’ve only ever seen cars and people in the driver’s seat, and it’s uncomfortable, and expecting people to just magically overcome that uncomfort even if you know you’re right, even if you know your technology works, even if you know that self driving car definitely can drive itself. People need help. They need help understanding why they should trust you, they need you to earn that trust. And that’s the most important thing that you can do. Because without that, what you’re gonna do is find people that reject it because it’s scary. It’s new is threatening my job, I don’t know. And by the way, if you don’t earn trust, and they do adopt it, for some reason, what happens the first time it’s wrong. Because if you don’t have trust, and it’s ever wrong, they will never trust you. You’ve lost, it’s done. And so building that trust moving up that trust ladder is the most important thing to do. And it starts with realizing that you have to be transparent. One of the biggest challenges we saw in the early days of outlier was that a tool like outliers was scarier data to find insights, there inevitably would be a person in the room who was afraid it would take their job or worse, it was afraid that everybody knew what was really going on. Since then they would lose their job for sure. And you can understand where that comes from. If you’re in a business where information is limited people can exist in the shadows, or they exist in through momentum. And there’s no accountability, because it’s so difficult to find that now you start to shine sunshine on it, it’s very difficult to get people to adjust to it. And so we realized early on that one of the most important things we needed to do was convert the skeptics, right? How do you take somebody who believes their job is on the line and move them forward, there’s one company we work with, the executives were very excited to adopt outlier, they kept missing things internally. So they tried it out. They loved it signed a contract, we came, we were set up for the first training with their team. And their team refused to show up because our team had talked and they decided that you’re going to use ally replace us. And we are not going to let you train it on our tool. So we refuse to show up and do it, they essentially had pseudo unionized. And of course, the executive team wasn’t having that and, and we got deployed at that company. But that anxiety that they had was real, and it’s out there. And the biggest challenge for any new technology, including AI is how do you overcome that? Because just telling people to suck it up and deal with it is not going to work? And so how do you prove the people this will make them better at their job, instead of this will take your job? How do you prove to them that this will help people make better decisions rather than have everybody criticized, things that this other person is doing. And as a result, I will tell you successful AI products have at least as much psychology built into them as there is machine learning. Because the machine learning itself is cold, and calculating, and it works. But if you don’t wrap that in the psychology of adoption, the psychology of understanding the empathy for these people who have had this job for 20 years, this is what they know. And all of a sudden, you’re coming in and threatening their worldview of how things should work. You have to build that into your product, because it is impossible to commit something otherwise,

Alexander Ferguson 32:02
thinking back to those first few beta test cars he went out to and just give us your data. And you said you found things. Do you remember what you found? Like, what were some of those insights?

Sean Byrnes 32:13
Oh, this was the crazy thing is we would go into these companies for like three days, we put out their data and give them a report on what we found. And we would, we would change the course of how they run their business. And just those three days. And it turned out it wasn’t because we were so brilliant. It was because there was just so many places that they just weren’t looking. And this is the kind of metaphor that I’ll use for you. Like imagine you everybody has a closet, or even maybe your garage or your basement where there’s just a lot of stuff piled everywhere, right. And it’s just, it’s so hard to find anything and you know that there’s stuff in there. But you just don’t go there, you just close the door and pretend like it doesn’t exist, because the amount of effort to go in there and find something is so high. That was what had happened to enterprise data is these big companies had data, they knew all these things. But the amount of effort it would take to go sort through and sift through and look for these sorts of things was just so momentous, they just didn’t. And us going in and looking through a lot of why they were such values. In some cases, we were the first person to ever look through their customer support data, we were the first person to look through not just their website traffic, but also their fulfillment and their payment data. And so a lot of it was looking across all these datasets, not just a few that were easy, but all of them and finding things that really did move the needle. I mean, there were just there was companies where we identified that they had very significant customer churn problems and certain demographics, some cases there was buying habits, they didn’t know that a certain kind of buying behavior is driving up to 20% of their revenue, because they saw the revenue. And that was great, what they didn’t realize it was a very specific buying habit that was driving a lot of that growth. And because a lot of those things live in between, right it lives in between your your ads that you’re running, and your website or your website and fulfillment or fulfillment and customer support. And all these systems have been spread out. And by looking through them looking across them, all of a sudden you start to see these questions come into into clarity, because you’re not treating it as a bunch of pieces of you’re putting together for them. And that led to a lot of the inspiration for building the software is like That’s why if you were to pull if you were open the hood of outlier. It doesn’t look anything like a traditional BI system, it’d be like looking at a Tesla versus a combustion engine. It the approach is so different because what we realized that doing this as that approach had to be different, like traditional approaches just weren’t going to get where we need them to go. We need to rethink all of those things that people thought they knew about business intelligence to be able to build something that would do this. And that in and of itself was very liberating. Because once we started to realize that all of these things that we thought we knew where they were what exactly was holding us back and he throw those away. Now you’re like, okay, cool. If we were to start today, what would we do? Well, we’re not going to save your data’s in one place. Because it’s a dozen places. And by the way, let’s assume it’ll never be one place. And then your data is dirty people and data quality issues, let’s not assume those go away. Let’s assume you always have them, right? I mean, you start to you started to go through this, you’re like, okay, cool. Well, I’ll cross this board up designing a new approach, based on the constraints today, instead of what the constraints were 20 years ago. And that also, I think, was a big part of why we’re able to make it work and make the product what it is today.

Alexander Ferguson 35:34
What does the future look like then? With your with your perception of of the space, your understanding of technology? What is it? What does if you make a prediction, what does it look like?

Sean Byrnes 35:45
Well, I’ll tell you that I this category of software is going to overtake how people make decisions. And I’m not so egotistical to believe everyone in the future means outlier. But think about this way. It wasn’t that long ago, 1015 years that if you want to drive somewhere, you had paper maps, and that paper map would tell you where you were you try to find yourself and try to find a route, you get lost because maybe the map had mistakes. Or maybe there was traffic or construction. But it was literally the best we had for hundreds of years were these paper maps that we used to get around. And over time, we added more and more information, and they got denser and denser and harder to read. But they were the best we had. And all of a sudden, almost overnight, we had GPS on our phones, and you just told your phone where you’re going and knew where you weren’t knew where there’s traffic be where there was construction, and you focus on driving, and all of a sudden, nobody got lost anymore. And we all got one we’re going we knew what we were going to arrive. And that kind of transformation for navigation for transportation is what we’re going to see in data driven decision making companies. Because the power, the dashboards that we’re using, in the form, some companies have three dozen, four dozen, five dozen dashboards, those are like the paper maps, they were the best we had, we kept stuffing more information to them. This kind of approach of automatically look through data to find insights and bring them to you proactively is like that GPS navigation. So it will touch every business. And by the way, this isn’t just my opinion anymore. It was my opinion, six years ago, now it’s everybody’s opinion, you can see we’ll look at how transformational is. So in 10 years, every business will use something like what we do in this automated business analysis category. And that’s exciting. I think it’s really exciting to imagine how that will transform decision making in business where all of a sudden, all this data isn’t a burden anymore. All of a sudden, now it’s empowering you. It’s like giving decision makers an Iron Man suit and all of a sudden you can do so much more. And what will people do with that? How much farther can it drive business? It’ll be really exciting.

Alexander Ferguson 37:47
Thank you so much, Shawn, for sharing the journey that you’ve been on of both entrepreneurship but also finally coming back to using what you’ve learned machine learning and applying it to this fascinating problem and and the future that you’re painting that everyone can start to see. For those that want to learn more, you can head to outlier.ai Thanks again, Sean was great to have you on.

Sean Byrnes 38:10
Thanks for having me. It was fun.

Alexander Ferguson 38:12
We’ll see you all on the next episode of UpTech Report. Have you seen a company using AI machine learning or other technology to transform the way we live work and do business? Go to UpTech report.com and let us know