Erik Martinez: Hello and welcome to this episode of the digital velocity podcast. I'm Erik Martinez. And today I'm thrilled to welcome Cherian Koshy, a visionary in data mining and AI technology who has pioneered innovative solutions to help nonprofits find and connect with donors.

Today, Cherian and I will discuss how the techniques he has used to find and connect with donors can be applied to direct consumer retail. Cherian, welcome to the show.

Cherian Koshy: Hey Erik, thanks so much for having me. So excited to be here.

Erik Martinez: I am so thrilled. I've been looking forward to this conversation. And just so the [00:01:00] audience knows, Cherian and I met in this public speaking class and he has the funniest bit that you will ever need. So when that comes out on video, we need to share because it is hilarious.

Cherian Koshy: Oh, you are too kind, you are too kind. It's great to be in that class with you and learn a lot and I'm sure we're both better off for being part of it.

Erik Martinez: Yeah, it has been an amazing experience and, honestly, just the vocal warmups. Yes, folks, we do vocal warmups. Have actually really helped every single day. So, Cherian, would you mind giving us just a brief synopsis of your journey to this moment in time?

Cherian Koshy: Yeah, you bet. It's a really short story in the grand scheme of things. I was starting off in college looking for a part time job to work around my class schedule. And a friend of mine said, Hey, I'm working with this nonprofit group. Would you be interested? It's pretty good money, and I think you'd be good at it. And it turns out that you would be knocking on [00:02:00] doors asking people to support like the Sierra Club and other nonprofit organizations. I lived in Minnesota so, you're out like in the cold, knocking on doors, getting rejected 199 times out of 200. But it got me interested in what nonprofits do.

I ended up being pretty good at the job, better than some. And they asked me to go be the executive director out east. And I went through some training, spent some time out there, and that really solidified this is really interesting to me. This is a potential career. And that took me to some different places around the country and different roles.

In one of those roles, I was here in Des Moines working for a performing arts center during the pandemic. And, we had laid off a bunch of staff. We're trying to figure out how to stay connected with our donor community, our ticket buyers, our audiences. And we couldn't even go to the building. We're, stuck with not being able to hire consultants work with our vendors. And we're, [00:03:00] obviously down several staff members. I, at one point I joked that I had watched everything on Netflix and I had too many open bottles of wine, which is not far from the truth, but, I was not being productive with my time.

I didn't know what I was going to do. And this is about the time that OpenAI made their API available. And I was like, oh, this is interesting. Maybe I can play with this. And I started building stuff and it was the pandemic. So, like everybody was willing to help. I had y combinators on my speed dial on my cell phone. And they would hop on zoom calls to help me out. And they're like, what are you trying to build? Oh, that's cool. Let me suggest this and do this. So, I built this tool, and long story short I made the platform, I really built it for myself and our own team. But I made the platform public and about a year and a half later, we had over 200 customers from all across the world.

And I took it out really to look for partnerships with other companies in the space. And one company reached out and said, Hey we're not interested in partnering, [00:04:00] we're interested in buying your software. I didn't have that on the bingo list for the year. So that deal closed, and they bought me along with that.

So, I had to leave my nonprofit job and I, they gave me a job as a VP at this company working with the product roadmap, integrating AI into some of their other tools. And now we're bringing even additional product features to market as part of that, but that's the experience that I had, and now it's really just a joy to be able to go out and speak to audiences, sometimes about AI, but sometimes about other things, and be able to meet people all over the world, which is really cool.

Erik Martinez: Yeah, that sounds really cool. Wow. A year and a half. And then you sold your company. Impressive. I think a lot of people that are listening audience would love that to happen.

Cherian Koshy: Let's be clear, Erik. It was very much first to market. It wasn't the most innovative thing. Wasn't the most technologically coolest thing. It was solving a use case problem for an [00:05:00] audience. And certainly, a widespread audience. There was proof of product market fit. There was revenue not a huge ton, but there was some revenue, but really it was first to market. Was the first company to be able to do some of the things that others were trying to look at. How do I do this and how do I do this differently? And I would still say that the. The product that I built and now has been enhanced several fold with where it is now is very far beyond what other people in the market have made in the same sector, but it's not so much as the technological expertise, as much as it's like right time, right, situation.

Erik Martinez: But from a marketing standpoint, I mean, that's one of the things you just said was, I identified something that the audience needed. You identified something that the market needed. And, I think there's a lesson in that, like for all of our customers, like some of us have more, specific product, more unique product, although I would argue [00:06:00] today that it's hard to find anything that's truly unique in the world of Amazon and Wayfair and all these other marketplaces, right.

But your approach to the market can be. Incredibly unique and incredibly relevant. Let's pivot from that and let's talk about this article. I was up on your website and researching for this episode, and I found this article that you had written about how nonprofits can save time and money with AI led production in the areas of generating content, optimizing outputs and automating analysis. And I'm sitting here thinking, gosh, that's what we do every day. So, could you tell us a little bit about what you learned by implementing those principles with your clients?

Cherian Koshy: Yeah, so, the first thing is everybody's trying to get in front of their ICP, their ideal customer profile. And when we're specifically talking about generating content, the [00:07:00] barrier is the blank screen syndrome. Where do I start? Because it's always something that you feel like you can do later. And there's never enough time in the day to block out time, and I'm going to be, in this moment, start writing. Whether it's a blog post, or a script for a YouTube video, or an Instagram reel, or whatever it is. I'm terrible at social media. But you know, whatever the content is that your company is producing, what I remind audiences all the time is that the goal is not to have AI automate everything and put it into production to the point of posting it on your website, but to really create that first draft so that you don't have that blank screen syndrome.

A lot of people think about like ChatGPT and writing a blog post or a LinkedIn post or whatever. That's a great starting point. I don't think it should end there, but if you start there, you have this content that you're reading right now started from an AI tool that said, this is who your customer [00:08:00] avatar is. Your audience avatar is here's your brand voice. Here's what we suggest you improve the title with. And here's the content that they might be looking for. So, I got an outline and some talking points. And then I was able to go in and add some context, some examples that were relevant to my experience with clients or organizations I've spoken to. So, the real key is just being able to produce the content faster from idea to implementation, but also to talk about how to vary that content for different audiences. So, if you create two different products, and you're sending out emails about both of those products. You can probably guess that people care more about one than the other, but you're too time strapped to create two different emails. And if you can create those different emails meaningfully that would apply to those specific [00:09:00] audiences, you're going to get more connection because they feel seen, they feel heard, they're more likely to convert.

So those are the types of things that I talk about. Certainly, with nonprofit audiences, if you're an animal shelter, talk to the cat people and show them pictures of cats, then say we serve cats and dogs. And talk to the dog people differently show them dog pictures, and we help dogs, right? That resonates with those audiences in a way that makes it land with them better.

Erik Martinez: Yeah, I think one of the challenges though, just using that example, right? Because in the direct-to-consumer world that. We operate in, what you're talking about is basic segmentation

Cherian Koshy: Yeah,

Erik Martinez: And we understand it very well. The challenge that I think we all run into is, hey, I've got dog people and I've got cat people, Oh man, now I have to create two different lines of content. And I'm a personal believer that the [00:10:00] hangup for most of us is. We're struggling with that. I really have five great segments that I can talk to and reach, but I have time to produce one. So, how do we take AI tools and do that a little bit quicker so I can reach the dog people and the cat people. And maybe the gerbil people. How do we do that so we can scale quicker and have that great connection with the audience?

Cherian Koshy: Absolutely. So, let's take a very specific example, like a triggered set of emails that you might send to cat buyers versus dog buyers. And we could even get more complicated. With prospective buyers versus repeat buyers. Like you're going to upsell on the cart or something like that, but we don't need to do that in this moment.

We can tell the AI tool let's use chat GPT for this moment, because everybody's super familiar with it and say, here's what my company does. Here are the products and services that we offer. You can even say, [00:11:00] read this website and here's the content or here's our product brochure. Now create for me. Two customer journeys for cat people and dog people and a five-email drip sequence or a five-email trigger to email sequence for each of these segments. I would suggest just saying to keep it clean. If you have five segments, let's start with the first one, run all of those out, move those into your email automation tool, then do the next segment, do that over and over again, so that you can make sure that it's the right thing that you've edited it, and that it's what you want to accomplish.

Cause you can go in there and say, Oh, I don't like email number two, fix that. But at this point you're running. Like days worth of work in minutes. The thing that people get overwhelmed by is they're like, well, I have to do all this prompt engineering, and I need to figure out blah, blah, blah. And that's all fine and good to a certain [00:12:00] extent.

But remember, it's in the name, it's chat GPT. You're having a conversation with it. So rather than trying to get precision on the first attempt. Fire than aim. Don't overwhelm yourself, just say, I want to write this, I want to write one brochure for cat people and one brochure for dog people. And it will do that.

It will understand at a general level what you're asking it for, but it's not going to read your mind. So, there may be something, oh, well, I need to talk to prospective cat buyers. not existing cat buyers. So, you say, okay, fix this for perspective cat buyers. And it will do that. And, then you keep asking it questions and prove it this way.

And then you move it into your automation tool, or you move it into your blog infrastructure or whatever. But now, in addition to the five segments that you have, you can create an infinite number of other variations, right? So if [00:13:00] you have a cart abandonment sequence, you could specify that to people who looked at cat toys and write content that's specific to, hey, I don't know what a cat name is.

I don't, not a cat person. So, let's make it dog. So, it's easier for me. If they looked at dog toys and they abandoned that cart and you know that, and you want to do it or an abandonment email to them, you could be specific, pull this piece about dog toys and use that to write a cart abandonment email and it will draft something for you.

Is it going to be absolutely perfect? Maybe not, but you've got a draft to work with. And that's really what I think the goal is to be able to put into your sequence.

Erik Martinez: What do you say to people? Cause I had this conversation with a client fairly recently where. We were working on some SEO optimization and just trying to get the, they've got 50, 000 [00:14:00] products, so it's a heck of a lot to optimize and some of the product is fairly similar they sell auto parts for some classic cars and, each model has a unique set of products. And then there is the common set of products. And so, we generated a couple of prompts to generate, the one-on-one SEO stuff, product name, product title, page description, things like that. And I gave them two different options. And I said, here's the thing. This was a sample of a thousand items, and I can tell you just looking through this really quickly. We need some of your guidance on how to fine tune this because. it's still too AI, if that makes sense.

Cherian Koshy: Oh, sure. Huh.

Erik Martinez: And we need to refine it a little bit with some of your expertise. So, it needs a little editorial and they shot back and like, [00:15:00] we've got 40, 000 products, Erik, I don't want to editorial.

Cherian Koshy: Huh.

Erik Martinez: So, how do you address that particular issue?

Cherian Koshy: So, with a product listing that size, there's a few ways to go. There's a tool called make. com that would allow you to run prompts against a list, essentially. So, it's like Zapier with ChatGPT built in. You can make ChatGPT one of the connections. So, you would just pull the list of products and then say for each one of these products, here's the prompt. And what I would suggest is taking a brand voice. For your auto parts company like how they like to talk, how they want to approach it. That's going to get really close to what they would be comfortable with.

And then your SEO optimization and what it will do is it'll run through a Google sheet or an Excel spreadsheet for each one of the products, and it will complete the prompt line by line. That's really cool to watch actually. And it will produce SEO for each one [00:16:00] of those things. And it costs like 11 bucks a month to do something like that. It's not expensive, but it's just going to, plus your chat GPT costs, which is 20 bucks a month, but it'll just run that over and over again. And now it's all done, right? All 40, 50, 000 products are done for you. There's no way to eliminate the human oversight piece, right?

No matter what you do, eventually you put 40, 50, 000 products on a website. Somebody's got to look at it, and say, this is okay to post online. So there's that piece. The other way of kind of thinking about it is what are the places where SEO optimization matters most? Does it matter at each one of the product line, product levels or does it matter at like category levels? Those are things that you're way more of an expert than I am., I don't know that stuff at all, but do you have to actually have optimization at every product? Some bolt for a classic car or is it fine to just be [00:17:00] like. I'm searching for a 1968 Chevy. This is what needs to show up.

Erik Martinez: Yeah, no, I think you bring up two good points there, right? We don't have to optimize every single product on our websites and yet you do,

Cherian Koshy: Okay.

Erik Martinez: Right? So, it depends, yeah, nobody cares about the nuts and the bolts in this particular case. But if I'm a women's apparel company and I'm selling a blouse with four sizes and four different colors. Color and size are hugely important in that particular instance.

Right? So, what you've given us is a Hey, here's a way to get all of that done really quickly. But more importantly, you said we can't get away from the human element.

Cherian Koshy: Yeah, for sure.

Erik Martinez: It'll reduce it though. Cause that used to take. How many human hours, to generate all that? And now you're just talking about editorial, which is a lot easier in the process. [00:18:00] So that's generating content. Your next piece in that article was a little bit about optimizing outputs and automating analysis. Let's talk a little bit about optimizing outputs. What did you mean by that?

Cherian Koshy: So, what I meant by that was around how a lot of organizations that I've worked with, a lot of organizations I see do the spray and pray approach around their content. So if it doesn't, specifically land with a particular audience there isn't a lot of work that's done to optimize.

And there's a few ways I've worked with organizations to figure this out. One is just basic preemptive testing, which you could do through artificial intelligence now because you can run simultaneous tests against mirrored buyers. You don't have to look at your existing buyers, artificial intelligence can pull a list of dog toy buyers in Michigan and run fourteen different ads against it [00:19:00] and see, without you losing any business, without you worrying about exposing your brand, it can say, which of these actually leads to conversions before you do your A B test on, these are the ones that I want to try with my customers.

That's where I think it's the combination of predictive and generative AI. Especially when it comes to social media ads Google ads, those types of things that cost you money. It's an insurance policy against your spend. Spend the time and the money doing the tests in advance in relevant way, get to the versions that are most likely to be successful. And then you're optimized as best as you can prior to the test. And now you're running a test with two really valid options, as opposed to just I think these two ideas would work, but you have no idea that there was an option C that you haven't tested at all. And you're not leaving money on the table with real customers.

That's the, I think the big [00:20:00] deal.

Erik Martinez: So, I'm going to put my tactical hat on for a moment.

Cherian Koshy: Yeah.

Erik Martinez: Cause strategically that makes great sense. Let's test our ad copy and creative before we deploy it, to give ourselves the best chance of success. But tactically, are we talking about using AI to create synthetic audiences?

Cherian Koshy: I am not in that instance. What I would say is you would want real audiences, real people who are actually making an evaluation of whether to buy. in this example, but you are using AI to create different ad variations. So you take your social media ad, your Facebook ad, for example, and you create three, four or five different versions or 15, it doesn't matter, like it takes 10 seconds to create 15 versions of this ad, but have meaningful differences, like different CTAs, different pictures, different whatever it might be that you think might be relevant or, an agency tells you these are the things that you should test.

And then you run them against real [00:21:00] buyers, and at least you have some data that these are the ones that are most likely to succeed. Let's narrow this down to the best performing two, three, whatever. Or maybe you end up with four or five that are similar, and then you take the five and you tweak those a little better in the test audience and now you end up with two or three, and now you can run a real test with actual buyers optimizing your spend.

Erik Martinez: Yeah. So let me translate that real quick. Cause I think what you're saying is, hey, you're going to use real buyers, real spend. So, you're going to take a small portion of your budget, which is something I advise all my clients to do is whatever we do from a spend standpoint, let's have a test budget to try things.

Because if we don't know how to learn we don't know how to improve. So, take that test budget, create some variations using AI to help you figure out what those are and then put it in the field and get some data to help optimize. I [00:22:00] think that's, there is a group of people out there like, hey, AI is going to do everything. I just hit a button and it's going to spit something out. And then I think there's another group of people who are like, Hey, I don't know. I think it's scary. And I don't even really know where to start outside of doing a real simple, help me with my email prompt when I respond to my boss on this particular topic or whatever, right?

And the reality is what you're saying is there's a really practical way of approaching AI, which is just saying, one, it's a conversation. You could talk to it like another human being, and I think we're so geared to the Google world of search, refine, search, refine, search, refine. Yeah, I'm wired that way, I grew up with these tools, but what you're saying is throw the idea out there and see what AI comes up with and then ask another question or two or three or four until you narrow in on what you're talking about. I think that's [00:23:00] brilliant.

Let's pivot now to automating analysis, because I think this part right here, when I was reading it, I'm like. Yeah, we should be doing more of that. And you talk about this concept of attention mapping

Gosh, I can tell you, study after study, we just completed our own behavioral study and conclusion, the buyer journey is a mess. Like there's a group of people out there who are touching every single piece of technology where they can consume content. And there's a group of people out there who do not want to touch any piece of technology. And then there's the group in the middle of who are like, Hey, we're using technology, but we're really busy, we can't do everything. So, that's a wide variation. What was really interesting in the study is it's like, it's a third, a third and third.

I'm like (explosion sound),

Cherian Koshy: Yeah.

Erik Martinez: And, by the way, demographically these people were evenly distributed. Just that last group, the one that was little [00:24:00] technology adverse was a little bit older than the rest of the groups, but not dramatically so. There were people in their twenties in that group.

Cherian Koshy: That's fascinating.

Erik Martinez: It is fascinating.

It's not what you would expect. And so what the implications are in any of one of our customer profiles or, in the nonprofit space, it would be even amongst your donors. You're going to have those people. A third of them are going to be like, yeah, we consume content everywhere. There's a group that's going; we only consume content on two or three things because that's all we have time for. And then the group that's like, I don't really want to consume content this way, but you're forcing me to write. So, in anybody's business, they're going to have a group of those people.

Now it may not be a third. But the idea is, if you want to reach your entire audience. That has huge implications for your marketing.

Cherian Koshy: For sure.

Erik Martinez: So back to the attention piece though. Regardless, if you're doing most of your stuff offline or you're doing most of your [00:25:00] stuff online, or you're doing some combination of both, think about all the different touch points that you can reach a customer and then suddenly you go, (explosion sound)

Cherian Koshy: Yeah

Erik Martinez: So how do we do this attention mapping thing? And how do we automate the analysis around it?

Cherian Koshy: So, I think that's the issue is like we have too many data points. And because there's so much stuff coming at us, it's really hard for us to see the forest for the trees, right? There are some things that we're trained to look at, and we're looking at conversions and whatever. But your preferences, whatever you look at, it's totally fine. But the challenge then is that you have attention bias. Because you're looking for certain things in the data that match up to what you most care about what you most want to see. And it's not to say that you're changing the data to meet your preference, but you're focused on finding those pieces in the data.

So, Whether it's, a free and expensive tool like chat GPT or some of the more bespoke tools out there, [00:26:00] there's a couple like Accio. com or Ali that do the sophisticated analyses and they're not coming at it from a particular lens. They're just taking all of your data, dumping it into one space and then surfacing up insight. That maybe is helpful to you. Maybe it's not. And then you're starting to train it and say, okay, I want to understand. What is that third, a third for my customers? Does it fall this way? What is the demographic piece of this third? What are some unique elements of this buyer segment? What might resonate with that group. And as you can imagine, now we're combining all of these other pieces that we've talked about together because now we have these insights that are being surfaced up for us. We're figuring out what those pieces look like. And just as you mentioned before, we can automate some of these tasks.

Chat GPT came up with this task thing, like every morning it will do [00:27:00] this this a little bit more complex to be able to do something like that. But if you're surfacing up and running this work every day, every week, every month, whatever, and that's driving particular strategic decisions that are then pulling together test ideas or test variations and that's driving content creation that are examples and then all of these working together in a harmonious cycle to give you more leverage will now all of your systems are working better for you, but you don't have to go through the manual task of analyzing the data, what used to be really expensive work to figure out what is the predictive model that would help us understand why someone is going to convert on this versus abandoned on that.

Those are all the pieces that could surface as part of this. The other thing that I raised in the article is we have some really sophisticated tools out there now that can [00:28:00] do things like eye-tracking technology. You can do that in real life - retailers do that, where they put eye-tracking technology on willing subjects, obviously, and then watch where they look inside of a store. What are they inclined to look at? How does that drive where product placement is? But certainly, on things like websites, that has huge implications than on website design - on where you're putting certain pieces what's driving those behaviors. So, once we get a sense of how our customers are actually behaving, we have such greater insight into their ultimate buying behavior - their ultimate buying decision-making process.

And so, while there's a lot that we can do before that, at the end of the day, we have to observe what they're actually doing that lends themselves to a sale so that we can leverage that repeatedly to get to even more sales and continue to optimize. That's where I think it's been [00:29:00] hard in the past to map that out - to understand what's driving that. And AI is allowing us to get much further, much faster, as part of that process.

Erik Martinez: Yeah, I think that's incredible because you're right. There are so many tools available to us today, but I think it does start getting overwhelming again, right? Most marketing teams, like the ones we worked with on the general basis are relatively small, but even in larger organizations where they have more specialized departments.

Those teams are still relatively small too for the mission that they're trying to accomplish. So, I would say that the larger organizations out there in the world have bigger budgets and they can throw more people at some of these problems where the small, midsize market. And you probably see this in your nonprofits too from big to small, right? The big ones can dedicate a little bit.

Maybe they don't have all the resources they need either, but they can make a little more progress and headway in some areas because they [00:30:00] do have larger staff and larger budgets where you're small and midsize nonprofits same thing in the commerce world. There's one or two people running the whole ship. That brings this challenge of integrating new technology and figuring out how to start and what challenges do you need to overcome? So, if you were to advise somebody, one of your clients today, he was like Cherian, I get all that but I've got 50, 000 things on my plate and while this will save me time, it may not save me time today or tomorrow because I'm still learning.

So, what's your advice to that person on how to. Hey, let's start incorporating some of this technology. Then you already gave us some great examples, but now I'm still overwhelmed because you gave me like 10 examples, Cherian, and my head is about to explode.

Cherian Koshy: I get it. So, I get this question a lot. And I mean, what I appreciate most about being on your podcast is a lot of my work, even though I work for a for [00:31:00] profit company, a lot of my work in the past has been with nonprofits where budgets are super tight, where staffing is super limited. And. Where e comm has a bit of an advantage, right?

Like there's at least a little bit of breathing room, maybe more than a nonprofit environment. Both entities still have the same challenge of there's so much happening. There's a new tool every day. How do I navigate all the different things that are happening in all the different spaces and where do I get started?

The first thing that I would say is while there are a lot of new tools that are coming out in the marketplace. There are tools that are built into everything that you're using already that you should just pay attention to. So inside of whatever email platform you're in, there are already tools that are there. Google or Microsoft Office or whatever. There are tools that can help you with some of the work that you're doing. But before we get to your CRM, all of those different things. But before we get to the tools, the thing that people [00:32:00] focus on too much, in my opinion, and it doesn't matter the size of the company, or the staffing is the tactical piece.

They focus on what do I do, opposed to what is the problem that I'm trying to solve? And want to reassure everybody that's listening here that There are plenty of problems to be solved, but the technology shouldn't lead the work that you're doing. It shouldn't be the driver of the decision. Just because the technology does a cool thing doesn't mean that you should go buy the technology or invest in the technology and change your process. The most important thing is to step back and say, What are the things that are taking up a bunch of my time? And that absolutely have to be done in order for my business to grow.

Typically, when I work with a company, I will sit down and, in the workshop, we'll do this with the leadership. We'll say, okay, what are the key items that are either hampering our growth or that would. Really [00:33:00] help us scale our operations. What are the things that would allow us to win? And what are the things that are super repetitive, super data entry driven that are not actually helping either our staff or our customers, once we start to unpack those things, now we start to be able to see what are the processes that we need to fix? And then we can start aligning problem process to tool and allow the tool to work back upstream to address those other things, those other issues.

Cherian Koshy: There are plenty of people who are going to come at you and sell you on, even free chat GPT. But I know plenty of people who are frustrated with chat GPT because they're like, I don't know where to get started. I don't know what to do with this because they haven't defined. what it's actually solving. They're like, Oh, this is cool for writing a poem in Shakespearean language. Well, what am I ever going to use that for? I happen to use ChatGPT every single day, and Claude and some other tools. [00:34:00] And so when I'm seeing challenges in our workforce and the things that our people are trying to do. Customer success, for example, are like, this isn't working. I'm like, Oh, I see this problem. I know what tools to use in order to address this problem, to create those solutions for our customer success team or for our sales team. And we can put those pieces together.

So, my biggest recommendation for anybody who's listening is put all of the tools. on a shelf for now. Look at the work that you're doing, understand this work. And what is the outcome of that work that's needed? Because there's times where you're working on this thing. And we don't actually have to do this thing. Doing this thing, accomplishing this task doesn't actually help our business grow. And that's a bigger decision. That's a more important decision. And if we can pull that out and say, listen, we're spending too much time on things that don't matter. We've won already to a certain extent, and maybe there's no tool that's [00:35:00] needed in our business. But if you do identify some problem where you're like, this is super urgent, it's mission critical, it will drive business success. It will help our customers, help our staff, all of those things. But we just don't have enough time to do it. It takes too much person power manual entry, whatever it is. Now let's look at what does that workflow process look like? And then what are the tools out there that can help us resolve those issues?

Erik Martinez: I think I couldn't have said it better myself.

Cherian Koshy: Oh, you sure could have

Erik Martinez: No, I don't think so. I've seen you speak. You're good. No, I think that's right. We've got to start with what is the core problem or the mission at the end of the day? What is it? We're really trying to accomplish and then fill in the infrastructure from there.

Cherian, this has been fantastic. As we move to close out, do you have any final thoughts or things you would like the audience to know?

Cherian Koshy: Just that, I think for a lot of people AI seems like the shiny new object. [00:36:00] And I want to remind people that some version of AI has been around for a real long time, and AI is going to get embedded in a lot of the things that you use every day. So don't get stuck. Thinking about AI is a thing over here that you need to become an expert in.

Nobody needs to learn a new language in their business. Maybe you should learn a new language as a thing that you do. Like, go learn Spanish because you want to learn Spanish. That's a good thing to do. But you don't need to learn a new language in your business just to feel like you're staying ahead of the trend.

Focus on your business, on the work that needs to be done, and then see how the technology can support your core business decisions.

Erik Martinez: That's perfect. Cherian, thank you so much for taking time to come on our show and share your insights and experience. If anybody wants to reach out to you, what's the best way to get in touch.

Cherian Koshy: I'm terrible at social media, but I'm mostly on LinkedIn. And my name is pretty [00:37:00] easy to remember. It's the stands out pretty easily. But I also have that website cheriankoshy.com. Those are the best ways.

Erik Martinez: Awesome. Thank you again for coming on. Folks that's it for this episode of the digital velocity podcast. I'm Erik Martinez and I wish you a great day.

[00:27:00] Thank you for listening. If you have enjoyed our show today, please tell a friend leave us a review and subscribe on your favorite podcast platform. Visit the digital velocity podcast .com website to send us your questions and topic suggestions. Be sure to join us again on the digital velocity podcast.