AI For Improved Customer Experiences – Techstrong AI Podcast EP31

Hello and welcome to the Techstrong AI Podcast. I'm Amanda Razani and with me today I am excited to have Fred Koopmans. He is the Chief Product Officer for Big Panda.

How are you doing today? Hi, Amanda. Good morning.

I'm doing great. Wonderful. Can you share what is Big Panda?

What services do you provide? Uh, uh, big Panda is an AIOps company. Um, and so kinda what that means is that we use data and AI to automate, uh, IT operations and service management.

So kind of there's the people that build applications and infrastructure, and there's the people that, you know, operate and service that those applications and infrastructure. So what does AIOps do? It, you know, helps identify incidents before they become incidents.

It helps identify, I risk changes, suggest things like, you know, this is the likely root cause, this is how you should fix it. Here's how to make sure it doesn't happen again and again and again. And so, like there's a whole army of people in large enterprises whose, whose job and responsibility set that is.

And well, AIOps says, Hey, let's bring data and AI to automate as much of that as we possibly can. Wonderful. Well, that's a good segue into our topic of the day then, which is how generative AI is impacting IT operations.

So from your perspective, what are some of the key impacts that you're seeing with generative AI as it relates to IT operations? Um, yeah, you know, there's a lot of things to, to kind of dive into there, but maybe we kinda step back for, for kind of a moment here. Generative AI is a disruptive wave.

And you know, I think myself, I've been in career in the industry maybe 22 years or so. Um, I put it in the context of other waves I participated in. So kind of straight outta school I worked on like mobile, like make the mobile internet work, make mobile data work.

And from there it was video and then data, and now ai. And I think it's one of these things when like you're at the beginning of a wave, you don't quite know it's a wave. Like you can kind of feel it, but you're sort of like, okay, what does it mean?

Do I take my product portfolio and I add a little bit? Is this a sprinkling, is this a rebuild? Uh, and I I imagine there's a lot of, uh, engineering teams, product portfolio owners that are like, are kind of questioning a a lot of that.

And it's tricky. It's not obvious what, what to go, uh, into. There's, there's so much possibility.

Like the, you know, I, I bet you a lot of people have experienced this, like what can be built in a hackathon in three days was the kind of thing the reach research department would've taken a year to do. Um, and frankly it'll be better. Uh, and, and certainly much higher loc velocity, you can get better feedback, et cetera.

So it's very, it's just fascinating. It's like, it, it's a fun time to be working in tech. It's a fun time to be owning a, a product portfolio.

Uh, a lot of responsibility that that comes with that. How do you thoughtfully introduce this technology without completely disrupting, you know, your existing business and alienating your customers and all of that. So there's, you know, it's, so there's a lot to to think through out there.

It is a lot to think through. And to that regard, what have you experienced as far as, is there any staff resistance and, and on the customer front customer resistance when it comes to incorporating ai? Um, yes.

You know, there's both resistance and, you know, angst and we're not going fast enough. So like, you know, you, you do your, your your biannual survey, half the comets are like, why are we investing in ai? The other half is like, why are we investing more in ai?

Like, you, you get this sort of like polarizing kind of view and obviously like, you know, that that's just sort of one, one perspective. Uh, I think one of my favorite quotes, uh, I was meeting with one of our customers, uh, our customers by the way, typically, you know, banks, stock exchanges, airlines, healthcare providers, you know, major online media, like large enterprises, pretty much any, anything with online services that would be bad if they went down. Like that's kind of a good customer for a big panda.

So I was meeting with a large bank, uh, back in June and you know, I asked for the customer for his roadmap, like, before I tell you my roadmap, what's your roadmap? And he starts talking, walking through a bunch of different things. And he noticed he didn't say anything about ai.

And so I called him on, it was like, Hey, what, what's your AI strategy? And he looks at me, he says, my AI strategy is to make sure that vendors don't double my price for no extra value. I was like, oh, okay.

So he's probably had like one too many AI roadmap. Like, this is gonna be amazing. Everybody's thinking they're gonna double their price kind of a thing.

And well like, look, I think he has since kind of warmed up to some of the things we're working on our AIOps co-pilot, just as an example. But it just sort of, you know, people have like, their, their, their defense up a little bit on the customer side when they hear like another, like, AI is gonna revolutionize everything that, you know, their eyes start to roll a little bit. Um, so yeah, there, there's, there's all kinds of things, uh, there internally there it's not so much resistance per se, it's sort of like, well, how do you do it?

Like, do you have a hackathon and you just sort of like spur a thousand ideas and you have like random acts of AI all over your product that it's hard to rationalize and make sense and build an intentional product strategy out of, like, that might be a good idea just to see what surfaces in, in a couple of days. But that's probably not the best idea as you bring things to market, if that makes sense. Yeah.

And that brings up another issue that being, having a good plan for why you're implementing ai, not just implementing it for implementation's sake. And it sounds like, uh, that was a concern of his too. Is there a real end goal if we're incorporating the ai, therefore they know if they've gotten a good return on investment?

Y Yes. And you know, maybe three years ago you could have some cool tech, cool demo people would just, here's my money. Um, it's not 2021 anymore, right?

In 2024, you need business value case ROI demonstrable, um, frameworks for, for that. You know, for anything above like, I don't know, 50 K or something like that, uh, you, you need to have that and it needs to translate something. So, you know, to your point, it's not just about the technology.

Um, and you know, I've, I've been reflecting on this, like I, I'm not positive, but I think when I graduated university, you know, came out with a engineering undergrad and a master's degree, I was pretty sure it was like 70% about the technology. So like the, the balance of power, let's say within a tech company, 70% technology and 30% like everything else, sales, marketing, product management, g and a, like all that kind of stuff. And I think I've subtracted like a couple percentage points from that every year of my career.

Like, I think it's like maybe 30% about the technology. Oh, gen AI is a huge sort of leap forward in what's possible that unlocks a lot of things. But you know, good old fashioned product management, persona identification, uh, business value assessments.

Like I got this cool Eck that is going to reduce your operational load, saving you certain amount of CapEx and opex. Like you gotta translate it into their language. 'cause your buyer like, okay, and by the way, my career, all career has been in enterprise.

So things could be wildly different in consumer spaces or in a space where like your buyer is an engineering team that buys the technology directly. But that's different. But for me, I've always sold to, you know, large enterprises, the IT side or something like that, uh, is a little bit different there.

You gotta translate it into business value that they can kind of take to their CFO and sort of say, we buy this software and here's the ROI we get and here's the payback and it's gotta be in business terms like gen ai, it's like, you know, the geeks get excited about that part, the economic buyer that that writes the check wants to see the ROI, you gotta model it out for 'em. Yeah. Every step in the process is very important and it comes down to good communication.

Uh, yes. And well, that's internal as well. You know, I think there's something in kind of go back to that, that thought of like, okay, when you have an established business of some form, whether you're at 50 million or you know, 500 million or, or or beyond revenue, you have some form of established business.

And it's been a little while since you've done a zero to one thing as a company. You're, you're, you're well past that and now you come in with some new technology and like, well how do you, how do you even get started on that? Is this just a couple of feature enhancements that you're gonna sprinkle around your product?

Maybe, and maybe that's fine, but it's also possible you're missing something. Uh, if that's the only approach that you take. It's also possible that if you kinda start with a blank sheet of paper and reimagine like, I was building this product space today or going after like broadly the same market, how would I build it and what would I do differently?

And that's not only a technical thing, it's sort of like, who would I hire differently? What does my org chart need to look like? Where are the zero to one people?

Do I have zero to one people in this organization? Are they working on this project? Because if, like, it's very difficult for too many people in an organization to have to scale and maintain, you know, an existing, well-established business with customers that have certain needs on an ongoing basis and invent something entirely new.

Uh, that's like the, you know, the one to two kind of, uh, uh, thing going from one product to, to multiple products. Um, so yeah, there's a lot to to to go in there and I think being thoughtful and intentional about what it is you're doing and then maybe also building some, some wins along the way. Right?

So the way the big pan started is we had a hackathon, we, you know, spurred a bunch of ideas and there was some really good ideas that came out of it. Um, not all of those got productized 'cause they didn't necessarily fit directly into the portfolio, although we're actually burning down that list. Uh, you know, a year later now we're, we're, we're fairly well along at some point we really gravitated on, you know, a whole like kind of module, a new module of the product that is targeting a certain persona with a set of common pain points.

And that's starting to sell, you know, quite well at this point. It's really a bundle of features though. But that led us to our next thing, which is like, okay, let's rethink things like generative ai, you know, with a, a interactive natural language user interface probably running in Slack or Microsoft teams with access to lots of data.

That's exactly the kinds of things we do, but it's a different orientation. And so we've actually built, uh, a co-pilot, uh, lots of people building co-pilots out there, but I actually think that's okay. I think that a, a highly specialized, uh, co-pilot, the specialized to a specific persona makes a lot of sense.

Like, if you think about it, our workforce is very specialized, you know, I don't just mean like legal and finance. I mean, within legal, you have privacy attorneys, uh, and you have like the, like a whole different variety of these things and they can sniff out an imposter in, in, in a half a second or some sort of generic kind of co-pilot in a half a second. They need to say, Hey, this is the data that needs to be trained on, this is the language that we use, these are our use cases.

If you can build a highly specialized copilot that's just right for kind of my set of use cases and needs, that's far more interesting than a generic, you know, company-wide kind of copilot. Um, so we're doing that for itop. So first line responders, you know, the incident management teams, those that kind of get in a war room when something is down, they have certain kinds of needs.

Um, there's pretty common from one enterprise to the next, and those are our users and there's, there's data lying about their organization already, both kind of real time data, historical data, pull all that in. We have the access to the right data and you've kind of programmed like, Hey, ai, I need you to think about and look and interpret this data the way my customer would, where my, my specific persona would. And I actually think that's really cool that there'll be like lots of co-pilots.

I don't know how it's all gonna work, I guess. Like there's gonna be like my co-pilot talks to your co-pilot and then they get together and they tell us what we need to go do next. Like, that'll be cool, you know, kind of collaboration amongst the co-pilots when that, when that comes out, I think that'll be the next issue of concern is figuring out how to maintain and keep track of all these connected co-pilots and programs, Right?

So again, it's not just about the tech, it's about, well, how do we govern this? How do we build trust around this? How do we control it?

I wanna harness the power, but uh, you know, it might be a regulated company. I need audit capabilities and stuff. So like, well, this is why enterprise companies exist, is to come in and take some powerful new technology, but then make it suitable for the enterprise where they can scale it and control it and test it and roll it out and audit and, you know, et cetera, et cetera.

So yeah, that, that is a hundred percent where the industry's covered. Yeah, definitely. Well, if there was one key takeaway you could leave our audience with today, what would that be?

Yeah, I, my my thought is like, decide when it's okay to do some incremental thinking for a while. Like, let me just take one foot in front of the other, like one step forward on generative AI and decide when and who, maybe it's just as important to kind of pull off and sort of say, I want you guys to start with a blank sheet of paper and rethink, um, how we would do this if we were starting from scratch today. What would we build?

And that's true both for builders, you know, people on the kind of the vendor side. And I also think it's true inside of, you know, a company, if you're consuming ai, there's sort of like, I have a bunch of tools. What if they had a little bit more AI in them?

That's cool. Same time. What does this, what's possible now that wasn't possible before?

What's this world gonna look like in three years or five years? And if you only do the incremental thinking, you're probably gonna miss or, or like miss it by several years, at least that longer term thing. All right.

Well, thank you for coming on our show and sharing your insights with us today. Uh, my pleasure. Mandos uh, delightful to meet with you and and thank you for inviting me.

All right. And thank you to our audience and stay tuned because there's more.