In this episode, we take a look at the question Can AI Be Used to Automate the Mundane?

Grant
Hey, everybody, welcome to another episode of ClickAI Radio. So welcome, I am excited to bring into the show today someone that it's taken me several times trying to get to this, busy busy guy. This is Evan Ryan. Evan, I hope I said your name right, right. Because I've said some people's names wrong. It's not Yvonne right or not right on? Is it Evan Ryan? Right?

Evan
It is Evan Ryan. Yeah, you got it in the right order. Most of the time, it's in the opposite order to flip it around.

Grant
Right. Um, founders CEO, King, of Teammate AI, right.

Evan
I just tried to go by founder. Okay.

Grant
Awesome. All right. Tell us a bit about yourself, Evan, what's brought you to this point. And anyway, thanks for joining the show. All right. Let me stop talking. Go ahead.

Evan
Well, Grant, thanks for having me. We were talking before we started recording, I just hit it off from the first second, at least in my opinion, I'm having a wonderful time. So really with me, I didn't really notice this. But throughout my entire childhood, I really had zero tolerance for boring stuff. I was somebody who just really likes doing fun things. And I like doing things that are that are entertaining, that are challenging, that are fascinating, that are motivating. But I don't like doing things that are boring. And I will basically do anything that I can to avoid doing boring stuff. And of course, now Hindsight is 2020. But that's sort of how I view a lot of my childhood. Well, in 2016, I went to a conference where they had this guy by the name of Jeremy Howard. And at the time, or a year or two prior, he was the number one ranked AI researcher in the world. He was talking about all this stuff that you could do with AI. And he, he mentioned that he had this company that could correctly diagnose cancer from MRIs and CT scans. Cool better than a team of board certified doctors. Oh, this isn't 2016. Yeah, so I was just about to graduate from college. I was thinking that oh my gosh, like, this is just the coolest thing I've ever heard. I basically went into AI the next day. Did you really? Wow. Yeah, it was it was really sort of powerful and transformative. Over the next couple of years of sort of iterating around well, what what do I really want to do inside of AI? You know, how do I want to define AI? First of all, there's sort of no definition that everybody can agree upon. I really found that, you know, I hate doing boring stuff. I bet a lot of other people hate doing boring stuff, too. Yeah, yeah. So and so now we all know the pattern there for sure. Yeah. Yeah. And what I've found is really interesting is, you know, one of the biggest conversation topics around AI right now is will AI take all of our jobs?

Grant
Oh, my goodness. Yeah. I hear that so many times. Yes.

Evan
I mean, there are like so many ways to answer that with the answer, in my opinion, being No. But what I found is really exciting is when I bring people into the creative process about how can they automate the mundane, the boring, repeatable on their days, they're so excited, like the first time in a while that they felt like, wow, the future is so much bigger than my past. So for the last several years, we've been automating, we've been automating tasks and automating processes for small and medium sized businesses.

Grant
So let me back up. So you hear this guy speak. You're like AI is the place to go. Your passion is let me take boring out of life. So you're leveraging AI to do that to remove the boring. I remember my dad is really what got me into the, into the technology industry. He was in it if you can believe it, right? So he was like one of the early early pioneers in some of this stuff. And he I remember him using the phrase Boolean shift a lot. He would always talk like, oh, technology is a Boolean shift. And I'm like, What are you talking about? What's a Boolean shift, you know? But this is kind of a Boolean shift with AI something, it's going to totally replace us. But in fact, I think it's a shift to the right, where it starts to automate more and more capabilities. It does require us to re retool. But that's good. Because every big shift, you know, throughout our whole economic history, right, as a country, and as innovation continues to grow, we have to keep shifting to the right terms of the skill sets that are important and critical. So I don't see it as, as this big day AI is gonna come take over, and we're all gonna sit around and become people like on E. What is it Wally, that movie Wally? Right, where they sit around doing nothing? Right, I don't think is that, but I do think it is a shift in terms of the capabilities that will bring and therefore we'll get skilled up in other areas. What are your thoughts on that?

Evan
Well, I have a question first before, before I answer that, do you have a memory of when you were a child, where you first saw technology, do something and you're like, holy, holy smokes.

Grant
I do it was that this will highlight the age difference between you and I, it was my, my dad worked in this. So there used to be a competitor to IBM. They're called Control Data Corp, and my dad worked for them. He'd worked on some of the big mainframe systems. And I remember one time, he took me into one of their big mainframe centers, right, and we walk into this room and the fans are running. And the big computers are, and the disk, you know, the tape drive is going, you know, like, just like you see on the old movies, right? And the big disk drives worrying. And my dad sits down and starts coding some things on it. And the computer starts interacting with me saying my name and helped me, you know, asking me questions. And I remember leaning up to and going, this is what I'm gonna do. I don't know everything that just happened there. But I want to be a part of this. So yes, that that was my technology origin story. Yeah.

Evan
Yeah. For me. So outside of seeing this, like this was years ago, when I was like a small child. I remember my best friend was sending out invitations to a fourth of July, a fourth of July party in the United States. And, and he mailmerge at like age eight, e mail merge, like 500 contacts. Oh, that's everything that is just insane. Like, as incredible as mail merging. And AJ at AJ. Also, like, what is this thing? mailmerge? Oh, my gosh. Yeah, it's so fun to like, kind of think about, you know, there's this like, one time where at some point, you're like, I wonder what else it can do? This is just the beginning. Yeah. But to answer your question, you know, will we will we all end up just kind of like sitting around or being artists all day? I don't think so. The example that I use is, you know, right now you and I are talking over zoom. It took me three clicks in order to get into our zoom meeting. Right? And so what if I went, what if we went back to the 1950s. And we said that in 70 years, you're going to be able to talk to virtually anybody in the world, you'll be able to see their face, the audio quality will be crystal clear. And I'll cost you essentially nothing. But all 1.5 million telephone switchboard, switchboard operators will lose their jobs.

Grant
Yeah, yeah, it's it's true. Yeah, that Yeah, right. Yeah, no, people like, like, There's no way.

Evan
Right, and, and so there's always been this sort of technological destruction that has taken place. Even back when we humans created the waterwheel. And now you didn't have as many people moving water to generate power generate electricity, or when we created ladders, and, you know, there are all sorts of different pieces of technology that we've had, over the last several 1000 years that have changed the way that we worked. This is just another one, I think it's a bigger shift than then some of the other ones. But, you know, the printing press did the same thing. There are all sorts of people that were just writing books, they were writing the Bible over and over and over and over again. Well, they lost their job pretty darn fast. Yet we have more people on the planet than ever before, and more people are employed than ever.

Grant
To do this. I remember one of the startups that I was involved with earlier in my career in Silicon Valley, we're trying to solve the problem of we want to be able to allow software engineers and designers to design systems remotely over the Internet, right. And so in order to do that, we have some some design tools, but we got to make that accessible and you need to be able to manipulate it remotely. Right. This is before there's a lot of big fat network pipes. You know, you know, everyone's got high bandwidth, right. So we were trying to figure out how to get this to work. We ended up getting it to work, but the cost was prohibitive. Right Just so dang grim. So when I think about the progress of technology, it's lots of trying over and over again until you get to this point where enough of the complexity can be abstracted away so that ultimately the end user sees this simplicity enough to say it's adoptable for both ease of use as well as cost perspective. I think we're right there. They I, I think we're doing similar thing, right to go through the same round, which is, can I take some techniques that are advanced that have some value, but you don't have to have massive data science background in order to get the benefits from that, and I see that shift taking place where more and more of it can then become accessible to others? As the cost drops dramatically?

Evan
I've seen similar things, I I couldn't agree more. You know, I think AI is sort of one of those things that for a decade or two was promised but under delivered, no, yeah. And now it's so it's like it's coming in with a vengeance. And the tools that are being made are super intuitive. The use cases that are being documented, and copied and repeated are ones that affect a lot of people. And I think people are really opening their minds to the idea that they don't have to do things the same way that they always have, just because that's how it's always been done. So I think it's kind of a combination of a huge mindset shift. But also just the tools are flat out better than they've ever been before. And they're finally usable to the point where you can automate things and you can create API's with button clicks instead of with raw code.

Grant
Yeah, yeah. Which is, which is fascinating. I think it's bringing an order of magnitude capability, and will continue to do that, to the kinds of problems that we can solve it. And it's just, I think, at the crux of it right in others is you look to the future, if you can continue to provide this time to this kind of computational power. And add that in the future to things like quantum, when we start blending both of those, I think the kinds of problems we can solve becomes quite large. But let me pull it back to tell me about the problems you're specifically solving with teammate AI. So So you got into this space, you're looking to go solve some problems with AI, you want to make things simple, what are you making simple?

Evan
You know, we really love to help entrepreneurial companies, and primarily companies where the executive team really wants to grow, they really would love to grow, they'd love to grow five or 10x, and the next decade or two decades, or even shorter, but they can't bear the idea of doing it while radically increasing their payroll. Right. And the most common complaint that we hear is not you know, I have the wrong team or my team doesn't do great work. It's that my team is underutilized because they have too much stuff on their plate. And so what we really try to automate the problems that we try to solve, are really helping these team members helping these employees identify what are the things that you do in a day that you hate the most, let's not automate the stuff that you love, let's automate the stuff that you hate. And then we'll figure out together how we can automate this so that you never have to deal with it again. And so we do that with with companies, basically, across every industry. The the question, the second question that I get the most often is, what industries does AI work in? And my answer is always well, you know, AI is a little like electricity, like, sure there's the electric company. But every every company was benefited. And every industry was benefited by electricity. And it's kind of a similar thing. Yeah, you know, there's sort of no gift that you can give people that's more valuable than their time, in my opinion.

Grant
What would be some use cases that you're applying it to? Is it? Is it things like bookkeeping, is it, you know, mundane, you know, administrative tasks? Or is it in other areas that you're applying AI?

Evan
So every company is a little different? I mean, of course, there's like bookkeeping and accounts receivable, and how can we send invoices out of our ERP smoother or submit website forms for invoicing smoother. But there's also a lot of reporting, sort of data collection, data crunching and then putting that into a human readable format. That way, people aren't in charge of trying to figure out what the data says. Instead, the computer tells you what the most important things are to look for. We use AI to write newspaper articles, write and publish newspaper articles for media outlets across the country. Regardless of your feelings of the media, it costs too much money in order to be able to write an article especially for info journalism, like what happened in the local sporting events. So we're doing that all across the country and all across the world, all the way to multi day processes where maybe We're using machine vision to look at images and see, well, are there any defects in the products that we're working with right now? If there are defects, what are the defects and let's report back to the supplier, let's report it back to whoever's responsible in order to be able to get that quality control, like up to speed and up to where we needed to be. So we do the boring in the mundane like accounts receivable, we also do the really sexy and complex machine vision. And we're reporting back with here. Here's the percentage of products that you shipped us, for example, that that were manufactured properly or are up to spec.

Grant
That's quite a, that's quite a wide range of use cases. That's, that's amazing. Are you building your machine vision work off of the Open CV material or your you did this all by hand?

Evan
We use Fast AI's library, which Fast AI was the not for profit that was founded by Jeremy Howard. So basically, Jeremy was telling me in this conference, you know, here's all the great things that you can do about AI, by the way, I have a free and open course. And we found that their library is just absolutely unbelievable. So we'll try as hard as we can to be able to, to be able to build it from scratch. And the reason for that we originally did not try to do that we originally tried to use a lot of the open source. Yeah. But the reason was really interesting. It was that, especially when, when the process that we were automating wasn't related to customer acquisition, lead generation. Yeah, what was happening a lot of times was our was our clients who would get this new capability called this AI that saving, you know, hours or days per week in some cases, and they'd say something to the effect of, you know, everybody else in our industry faces the same problem. Could we license this to everybody else in our industry? Wow. Well, now they had this old legacy business, that they flipped into a software as a service business. And we realized really fast that we that we had to be able to make sure that our solution scaled to more than just one user. Mm hmm.

Grant
Amazing. So what's been some of the outcomes, you've noticed now that you've been out doing this across a range of companies, what, what's been the impact to them?

Evan
What's most exciting for me is that all companies are unique, yet all companies are the same. So they all have product delivery, they all have accounting, and bookkeeping, they all have sales and customer acquisition, they all have customer service. And so they all have these sort of functions, that interplay really nicely together. But largely, they're the same. The real differentiator, among a lot of companies is things like their supply chain, their product and their product delivery. So being able to help a wide swath of companies get clarity surrounding, you know, AI isn't just for Silicon Valley. It's not just for Tesla, and Mehta and Apple and Google, right. That's been really exciting. I think that there's a real market shift going on among employees, I think employees really have a smaller and smaller tolerance for doing stuff that's boring and doesn't move the ball forward. And companies are really incentivized right now, to outsource a lot of the work to AI in order to be able to retain their best talent. And so that's been one of the really, really interesting things. I think that's come out of the last six months.

Grant
Now, I guess I could imagine it would be something like removing them in the mundane so that you can tap into their creative, right, I think that's really sort of what you're after, right? It's try to exploit opportunities to get more creativity from your people. I would imagine in today's market, too, with a lot of attrition taking place and the challenges with hiring, that this also can be beneficial. It's is it part of a play to help people stay in their jobs where you could take some of the mundane out of it, and therefore allow them to do more creative and enjoyable things in their work?

Evan
It absolutely is. That's actually one of the biggest reasons why a lot of companies decide to work with us right now is because they know that they if it can create a work experience, that's five times better than what it is right now. That key employee might not go looking for an extra $10,000 A year or an extra 20 grand, or for new opportunities just because it doesn't feel like a right fit anymore. And so we're seeing that all the time. We're also seeing on the flip side of it, a lot of companies that are having problems hiring, or they're having problems retaining employees, just overall, maybe they have a 3040 50% attrition rate on new hires in any particular role. They're starting to ask the question, Well, I wonder why. Like, maybe it's us. That's the problem. And it's not them. That's the problem. And so all of those tasks that used to be hired, are now being automated instead. That way they can hire for those more creative and fascinating and motivating roles. I mean, I don't know anybody I don't know anybody who wakes up on a Monday morning looking forward to doing the same stuff they've done for the last five years.

Grant
Yeah, if they do well, yeah, that's another conversation. But sounds like you use a range of things from RPA to some custom built AI work, you guys have developed is your sort of toolset is that, right?

Evan
Yeah, whatever it takes to get the job done. A lot of times customers will have specifications. But, you know, Microsoft Power automates a really powerful tool. And there's a lot that you can do with 1000 lines of Python code, right, and with a great AWS or Azure suite. And so we do use, we have a handy and a really kind of wide tool belt. But what we find is that we're using the same tool sort of over and over again, which is really handy, I think, overall, and it makes it so that the tools are getting better over time.

Grant
They are Yeah, they're definitely getting better. Okay, so for our audience, what would be a call to action for them? If they were wanting to learn more about you and your organization and what you're doing?

Evan
Yeah, so I think the biggest thing would be head to automation secrets that teammate ai.com, there, you can get a free copy of my new book, AI is your teammate. And really what it does is it kind of helps distill down what are the mindsets necessary in order to be able to use AI in your world, whether you're a business owner and entrepreneur and executive or an employee? And how can you make it happen? You know, there's a lot of sort of how to guides for how do you make automation happen, but I've tend to find that they're all either way too high level, like, AI can only be used for chatbots on your website, or cybersecurity, or they're literally showing you lines of code. And so how can you make it happen, no matter what your level of technical capabilities are? So I would head there, no matter what, get a teammate, or you can get it on Amazon or Barnes and Noble or anywhere where you find books.

Grant
Oh, that's cool. So when you think about a bell curve in terms of the amount of time that someone would commit to working with a group like yourself, what does that look like, you know, the bell curve is or the efforts or the projects you do with them? Is it a two week four week, eight week? What does that look like?

Evan
I you know, for for relatively small automations. If you're if you're using a tool, like Zapier, for example, that would be two to four weeks. So very, very short, you could wake up this month, you have a lot of stuff on your plate that you hate. And by next quarter, you could have saved 510 20 hours a week, depending on what your job description looks like, all the way to six months to a year depending on if you have a if you have a really complicated sort of project that you need to automate.

Grant
Gotcha. Gotcha. Very cool. Okay. What questions do you have for me, you said you wanted to ask me some questions before we started?

Evan
I do want to ask you, yeah, what do you see? You know, so you do it a different type of like a eyes and really what we specialize in? We don't try to do a lot of the predictive the predictive modeling. Yeah. What are you seeing in the marketplace? On the predictive side?

Grant
Yeah. So on the predictive side, I'm definitely working in that space. Not so much in the CV area, but more in terms of predictive analytics itself. So you know, taking things like oh, how can I? How can I address stockout? problems, right, my supply chain? Or, oh, what can I do to increase sales? That is probably the number one use case that I see. Right, which is, hey, we're just trying to grow the business? And what are the conditions that are driving the best sales situation? Or how can I take costs out of my business? So efficiency plays? That's probably the second sort of style of problems that organizations need to solve. And similar to what you're describing, in all cases, it's Can I do it with the same amount or fewer resources? Right, I can't be adding more resources to this. In most cases, there's this FOMO aspect, which is there's this fear of the unknown, what is it that the AI can see that I can't write, because lots of times our brains are wired to see only just a few factors or variables. And then once we get too many dimensions out, our brain sort of gives out AI really exploits that well. And so casting as wide a net as possible, that makes sense for that business outcome. You're trying to target where it sells or whatever, and letting the AI help you to see all of those all those far reaching variables and pulling that in and saying actually, it's, it's the combination of these other factors as well means that this is when your sales take place. If it's this salesperson, during this time of the month to this particular market. You know, when the weather is clear in San Diego, whatever it might be, right? Those conditions tend to drive higher sales, whatever the situation, it's that watching business owners have that aha moment to go. Oh, okay. And that's, that's real satisfying, because then they go in and start tweaking their business, right, just enough to say, hey, it was worth the effort to discover them. One more thing, while I'm monologuing on this, there's that part. And then there's the other part, which is I find that AI, it can bring so many predictive insights, that it cripples the organization, right? It comes back and says, here's all the drivers, and here's all the factors, but it gives you, you know, 20 of them, and you're like, oh, okay, what am I gonna do with 20? Right? How do I figure and so that's the other key part of what we do, which is, we then say, oh, let's prioritize these into a series of incremental steps that moves the organization one step at a time. Otherwise, people get changed fatigue, right, it's too much to keep trying to, you know, do it all at once. So we take the insights that are predictive, go after those that have the highest probability as it relates to the business outcome, and then just go do one or two of those, and then rebuild, because contacts, you know, business shit, and business drift occurs, data drift occurs. And so you then refactor the the model again and gives you fresh insights. So how's that for a long answer? That's what I'm saying?

Evan
Well, that's actually you kind of touched on one of the follow ups that I wanted to ask, which is, we spend a lot of our time with the end users with the end employees, it sounds to me like you spend a lot of your time with the C suite. Is that correct?

Grant
We do, but it depends on the organization. And who's been tasked, in many cases, will we start with a C suite. And I'll tell you why. One of the challenges, I believe, with a lot of the AI platforms today is is the over over focus on model accuracy, right getting a 90% accurate, now, don't get me wrong, the model has got to be, you know, really accurate, but when it's done outside of the context of your business operations, then it means I could end up producing an AI model that's so efficient, that my business is not actually able to deal with or handle it may be bringing me too many deals, such that it actually increases the cost of goods sold, that it actually ends up hurting my business. And so it really needs this combination of a sufficient, efficient model connected to what are my business costs, my operations, my you know, the the amount of resources I have available, and that's why it needs to go a step at a time, right, you just keep going one step at a time to improve or grow it. So sometimes it's with the data, people. But if you do it outside of the context of those business questions, then it tends not to be as effective on the ROI.

Evan
That's a that's one of the things I was gonna ask, are you seeing that there are sometimes negative consequences where the AI is so good? Yeah, you know, that people either you have a change management problem where people's preconceived notions of why things happened was actually incorrect. And now you have to retrain, or something like that, where, you know, the like what you said, the cost of goods rises so much, because it's so efficient at acquiring new customers or getting more sales, that that the business wasn't ready to scale to that level? Do you see that that happens more often than not? Or is that a sort of a corner case?

Grant
I don't know, I don't think it's corner case. It's, it's a fair, fair amount of the cases, though, enough to be a worry, right, that if I don't take change management into consideration, as I roll out AI, then then my probabilities of success dropped dramatically that just because I have the insights from Ai, in my opinion, is only 70% of the way there, then you got to get that, you know, last mile and and the last mile is the successful rollout and adoption of this, and sometimes it's a cultural thing you're running into, people are worried, oh, I'm gonna lose my job. Others are like, Oh, this is gonna change my job. And then others are, well, we embrace it. But now we run into a money problem. And the money problem is that our business operations can't handle this adjustment. Or maybe the AI got it wrong, and the business can't handle that adjustment either. Right? Doesn't mean that it's always right. And so in either case, it can have that financial impact. And if we're not, if we're not taming the AI enough in the context of business operations, then it ends up creating a problem. So there's several hurdles after you get just those those predictive insights.

Evan
Yeah, one of the things that's interesting about hearing hearing your world which your world is just is so radically different than mine, I mean, with us, we have a pretty set, you know, this set of criteria. We're going to automate this process Is this process we've mapped out exactly what the steps are in the process, and then we build a computer to do it. with you what I think is interesting is, I hear all the time, that this concept of what we want to use data to make better decisions. Oh, yeah. And, like, there, I always think, you know, there's part of that that's true. But there's also part of it, that's like, you are thinking that the human should be making a decision right now.

Grant
I like to view this more as augmented intelligence. I know we say AI. But I think a should be augmented. It's really the state of where the practice is, I think in AI, to say that we're going to give all decision making rights over to some AI model and just blindly trust that I think that's naive in today's AI. Now, you know, they're getting better and better. But I work a lot of organizations where the majority, the AI model is early. And so it's growing, the need for a lot more cognitive support from the humans, to ensure that this thing is naturally moving in a way that is reliable, and truly predictable. And otherwise, I think you could just hand it off and say, I give all all decisioning rights over to the AI, I think that's foolish, you have the ability and need the ability, even after you've deployed an AI model to come back and vote on the impact of that insight. And that's important, because we want to continue to refine the training and retraining of the AI. Hey, what you just shared with me that predictive insight actually didn't pan out, that guidance really needs to come back into the models.

Evan
Yeah, I think, you know, the AI at the end of the day, like the algorithms, they make a prediction and they make a recommendation, but they never, they never make a decision. Now, humans either a prediction or recommendation, the human needs to make a decision. So the AI can provide all sorts of information, and it can provide recommendations. But But yeah, I don't it's not ready yet to to just understand how the world works and understand where you're going, what your objectives are, and then just say, this is right. It's not it's not quite otherwise. I know. I know, a lot of senior managers who are going to have really bad days. AI can do that. So what are some of your what are some of your sweet spots? What are the things that were the projects where you know, you know, that you can hit it out of the park?

Grant
Yeah, it's, like I said, it's in medium sized organizations typically trying to solve, you know, a revenue sales problem, right? That's definitely a sweet spot and supply chain areas, right? That's where they're looking to say, Hey, I'm trying to make sure I can, can keep inventory coming in at the right pace or the right rate, which is a serious problem now. But you know, we've also seen it in even in the current talent management shortages that's going on, which is, can I use it to help me understand the probabilities of, you know, certain groups or individuals who are candidates for leaving early right, and the cost and impact to an organization when that happens? Those are the types of use cases where typically we get involved. Those are, those are great questions, for sure. Okay. All right. This has been awesome. Yeah, you haven't. Thank you.

Evan
This has been great Grant.

Grant
Yeah, thanks for your questions. Any final statement before we wrap up about Teammate AI?

Evan
You can grab AI as your teammate on Amazon or at automationsecretsteamai.com. But mostly, I just hope that everybody has a future that's far, far bigger than their past, and far better than their past. Thanks for having me, Grant. Really appreciate this one.

Grant
Thank you for your time, everybody. Thanks for joining another episode of clique AI radio. And until next time, get some Teammate AI.

Thank you for joining Grant on ClickAI Radio. Don't forget to subscribe and leave feedback. And remember to download your free ebook, visit ClickAIRadio.com now.