Becoming an AI-First Company: The Yum! Brands Transformation | Domo BUILD 2026

I'd like to welcome Ben McClave, who is a Principal Business Intelligence Architect at Yum Brands. He will share how one of the world's largest restaurant companies is making the shift to becoming an AI-first enterprise. You'll learn how Yum is rapidly testing new features, scaling AI adoption, and creating a repeatable model for continuous innovation. Ben will be joined by a couple of my favorite Domo Sapiens, Sharon Allard and Will West. Now we will welcome them to the stage and turn the time over to them. Thank you so much.

Hi, everyone. 500 apps from your announcement at Domopalooza, Will, it's just been an amazing experience with our customers. I can tell you that 500 apps started at 5:00 AM in the morning when Will called me and said, "I have a great idea."

Yeah, a little bit too much coffee that morning, and what a wonderful experience. We had 166 different use cases that came in from a lot of wonderful accounts, Ben, your team included, where there's just a hunger out there where everyone wants to dive into what feels like the hottest thing going on right now. Some of my favorite use cases were custom apps that we used to drive data science, melded together with sales teams to drive sales and help the teams prep on a Monday for how they should approach the market. And then Yum as an example, where we're looking at how to make it so that the team has less cross-departmental reliance and the ability to essentially in-house things and speed the organization up. Just the process of all of that was such a wonderful experience. Sharon, did you have any that stood out to you that you thought were your favorite?

Yeah, the one that stood out for me really was Ben's request. It was about doing rapid development and doing an AI-first transformation in his organization. He was kind enough to share his vision about being AI-first in his company, so that was absolutely my favorite. We asked him to come today to talk about that vision. So Ben, I'm going to turn it over to you. If you could tell us a little bit about how this got started and what your vision has been.

All right, absolutely. Thank you, and I appreciate the invite to come and talk with you all. When I heard that you all were planning to build 500 apps, it was like finding the magic lamp and rubbing it, and the genie comes out and grants your wishes. What do you do when the genie first comes out of the lamp? You ask for more wishes. That's what I did. Instead of saying, "Hey, I want this one app," how about if I get an app that builds apps so I can get infinite wishes?

I imagine that like most teams when they saw the 500 apps challenge, they were like, "Oh, I want a dashboard, I want a summary, and I want this thing that makes this task better." That honestly is a really sensible ask. That's something you can turn on on Monday morning, have something practical, built for a purpose, and put into practice today. But what we wanted was something more repeatable as a model—something that not only solves a need today but can continue to solve needs going forward. We needed something that lets people do something in a fraction of the time, and it has to work across all of our brands. We have KFC, Taco Bell, Habit Burger, and Pizza Hut. I need to be able to flip a switch and change the branding, the look, and the feel in an instant. That's something that Domo has to do, too.

So, we didn't set out necessarily to find creative uses of AI. We weren't a hammer looking for a nail. We set out to kill bottlenecks in our operation to become more efficient, and it turns out that AI is a great way to do that. As a leader in the BI space, for example, I have a bottleneck that I'm sure you all can relate to. It's like Bill Yost's take on "If You Give a Mouse a Cookie." As soon as you build the dashboard wireframe, an executive is going to ask for more features, and more features, and more features. When your development cycle is slow, that gives them more time to ideate, the scope can creep, and it kind of gets out of hand.

What we did is we didn't point the AI to the easy thing, like summing up a meeting. We pointed it to the hard thing: the design. We take the transcript from our call with the stakeholders where they describe what they want. We take the visuals that they reacted well to, the visuals that they didn't react well to, and we give the AI all that context. Then we have it set to the task of building the wireframe. It can turn it around so fast and from so many different angles that I was surprised at how good the outcome was. I can give it that prompt, all that context, and say, "Now you're a UX designer, critique this dashboard. Now you're a chief operator, critique this dashboard." Each iteration made it that much better. It's all done behind the scenes in a chat window. We can handle it asynchronously with the feedback, and by the time we get to the second meeting, instead of talking about concepts, we're getting sign-off on the final product.

That's where Domo really comes in and brings it to the ultimate last step for us because we can take React dashboards and go to the pro-code builder. We can take that wireframe and deploy it very quickly and easily relative to how we used to do it. That becomes our actual infinite wish: can we just take ourselves out of the process altogether? Can Domo build an app where a user goes in and describes what it is that they want to build, and Domo knows from the skills we bring how to do it the Yum way? It knows how to make it a KFC-branded app with KFC red, or Taco Bell purple, the layouts we prefer, and our design aesthetic. We bring our skills and our design system to Domo, and Domo does what it does best, surfacing it in an app with real data, governance, security, and all of that. None of this is new; you could have done this with humans 10 years ago. That's how we had to do it more than 10 years ago, but it would take a designer, a BI team, and engineers in a ton of long meetings with lots of iteration. Now, we can accelerate that. In essence, our one app is your genie in a bottle for all your BI needs.

Thank you, Ben. The call that I would make to Ben and the team is that the mentality you guys have in this moment, where everyone is looking around and trying to figure out how to use AI, is incredibly impressive. We're being asked to use AI, we're being compelled, but you and the team have it built in. It's a little bit different for you; you seem naturally driven into it. If you could just give us one hot tip for everyone on the call: how do you install that as an organization, or how do you pour gasoline on that fire? We can build custom apps, but the culture itself that you've built for your team internally is what impresses me the most and makes me excited to work with you every time we do.

Absolutely. I think when people see what it's capable of and how much time it can save—not just individually, but for the overall process—it becomes infectious. Levi on my team is brilliant; he's been with Domo forever and knows a ton about this. I got a message from him at 1:00 AM saying, "Hey, I just finished connecting this new wireframe to a Domo dataset." Another guy on the chat was like, "You and Ben need to go to bed earlier," because it's infectious as soon as you see what's possible. The best thing I can tell you to do is to just ask the AI itself how to do it. I started with the end in mind. I said, "I would love to be able to just do X, Y, or Z. Is there a way to do that?" The AI can tell you, create the plan, and walk you through it. I have it set up sessions. I told it, "Hey look, I have about an hour to dedicate a day, so let's create a work plan and chunk this off." Do the low-risk, high-reward stuff first. If I build an AI agent skill that can build a wireframe and I never get it past the wireframe stage, I've still saved a ton of time in my workflow. I give a better brief to my team, I get sign-off from stakeholders much earlier, and it doesn't have to go all the way to the point of connecting it to real data. But if it does, that's much the better, and I just keep chipping away at it an hour at a time. It teaches me as I go.

‍