Real-Time Fan Intelligence: ESPN's AI Alerting System | Domo BUILD 2026

Now I want to talk a little bit about activation and distribution of your data. And to do that I'm going to bring on a couple of our friends at ESPN. Doug Crayman and Jennifer Lean at ESPN are here to show us how they are creating an entirely new way to track and monitor real time fan intelligence. And they are doing this by having built what they call an AI driven alerting system. So please join me in welcoming both Doug and Jennifer and also a couple of our own Domo employees, some of my favorites, Hunter Nielsen and Gavin Fraser, who are going to walk through how they're changing the game with real time fan intelligence.

It's a pleasure to be here. Jen, Hunter and Gavin, we are a small group, yet we are mighty with regards to what we've been able to do with Domo, the AI space, and we want to share that with you because what was just discussed is the perfect example of what we do. I think the example that was just discussed, which was all about Formula One and the McLaren team. We are big believers of everything F1 as well, but in the world we live in with regards to ESPN fan support, the model is moving at the speed of live and our job is to be brief, bright and gone with regards to the customer support we need to do. And a big part of that, the most important part of that is the bright part. And the bright part relies on fan intelligence. So a lot of what we do is all about aggregating, if you will, all the data, the dark data that comes in from every single fan conversation in real time and actually turn that into actionable data at the speed of live in real time. And that's always been a challenge, but with AI now and with Domo, we've been able to do something that we think is just second to none and exceptional.

So you know, before I pass it off to Jen to talk a little bit more, we put out a goal where we have all of this data, thousands of contacts that come in across the Omni channel everyday. And when I say the Omni channel, these are contacts into customer service that come in via chat, phone, email, web form, SMS text, social media, you name it. We offer that different option to our fans and what we wanted to do was build a real time query and conversational AI agent that would connect to our Zendesk, which is ACRM, our Zendesk CRM tool, search all of our data sets and come back with intelligent responses for any questions we have of our fan contact data in real time and directly in our Domo instance. And that's really what we did. So with that, now I want to pass it to Jen real quickly, but we will ultimately go under the hood with Hunter and Gavin. We're able to build this incredible service. But Jen, let me pass it to you real quick for any comments as well.

Yes, thank you, Doug and thank you Domo for inviting us to be part of this Domo build session. As Doug mentioned, we move at the speed of live and when we have live events like we just finished and wrapped up the NBA Finals. What's important for us as the fan support team at ESPN is to meet the fans where and when they are in need of support, in need of assistance and guidance. And so an example that we've been able to apply this live alerts, a trigger is for the previous fifty-five games in the NBA Finals, which as a native New Yorker, I could focus more on the game and lean into Domo for those triggers. If there are purchase issues, streaming issues and our Zendesk tickets, our conversation with our fans, our subscribers are ingested within those intervals of time. So that we get alerted when we have an influx of certain issues that impact our fans. And that we need to be in the moment to identify those issues, address them, share that with our product, our engineering, our marketing teams. And so that we can inform our internal stakeholders when these moments arise and we are able to input whatever, you know, ten pole major sporting events that may be of interest to us to have these alerts enabled. You know, when, when we think about our sports calendar, we're going to have the US Open coming up Wimbledon. And these are events that we want to meet our fans and be there to support them in their time of need. And so this alert page has been fruitful and insightful on in, giving us that information in real time.

I also want to be I also want to share what we've also built with being able to report on the findings post event. So the live event data is for that's in the moment. But what happens when our leadership, our marketing team, they want to be able to have an idea, a summarization. I'll be able to have these takeaways of what has happened after an event has taken place. So we're able to identify and ask about known issues, whether it's within the fifteen minute interval, whether it's today or the week prior, we're able to ask the AI what are issues, what are takeaways, what are executive and leadership insights that we can share internally. Because with fan support, the buck does not stop with us. We are empowered to share information with our leaders and our executives so that we can make impactful data-driven decisions. And we can, you know, have an influence on our road map and how we prioritize the sports calendar moving forward.

And to add to that, I think something, you know, that matters most is we need to be able to get this data in the moment to our leadership. And it's customer service that we oversee. And, honestly, nobody reaches out to customer service to say that, you know, hey, everything's wonderful. That's really not what we're there for. What comes our way is often bad or ugly. And with that, whether it's anything impacting the product, the pricing, product promotions, interacting with the product, product limitations or overall application use, we need to have that information and never provide anecdotal responses. We need to be precise with what we're given. And with so many contacts, it would have taken days if not weeks to be able to produce those results for executive asks and summaries. But now with Domo AI, with the LLM use we have, it just takes seconds and we're able to provide the most elegant, comprehensive reply written in the most elegant form. And it also offers predictive information that allows us to move and create more compelling and better solutions going forward.

It's interesting. And I just want to add this before I hand this off a bit further when it comes to AI I saw something on the CNBC this morning I thought was really interesting. It was called the bullets versus butter dynamic with regards to economics and how actually now that's applying to AI. When nations and economics look at, you know, how do they manage that? You know, do you create a military infrastructure that's bullets and put all your money towards that? And thus then there's the butter, which is obviously feeding your people. There's that. How do you find the balance? What we're seeing in Business Today, especially with AI is a massive ask, a massive push to AI infrastructure, push to AI proof of concept with AI hyperscalers. Look at the different opportunities for services with, you name it, for business intelligence, for different use cases. But the Butters environment is really where Jen and I live, where we're able to provide an outcome of look at what the AI with Domo can actually do to provide, if you will, sustenance to our executive leadership that allows them to move more efficiently and create a more elegant product. That's what Domo has been able to build for us. So I'm getting ready to pass it off to Hunter and Gavin. But before I do, Jen, is there anything else you wanted to bring up before we make the leap to look under the hood to show all the exciting stuff we built with this AI prompt agent that you see here and the alert triggers?

Yeah. One thing I, I do want to close before mentioning it before we hand it off to Gavin and Hunter is that, you know, we're, we're, we're there in the moments with live events. We're there with the AI to, you know, provide insightful summaries for our leadership. And when we look into the future, we've also created the Super Bowl insights. Were we able to listen in to what our fans are talking about in terms of the Super Bowl and be able to prepare for next year when we host it, Doug, and we'll be able to empower ourselves with data in order to meet the fans where their concerns are. And Domo has empowered us with the ability to have more complete, actionable insights in the products that matter to us and our fans.

Absolutely. And again, this real quick is we're listening to the fans on their terms. So we've opened up our listening, if you will, in real time to any inquiries regarding the Super Bowl super early. I mean, there's really not any talk about the Super Bowl at this point in time. Season hasn't begun, preseason hasn't begun. But as soon as conversation starts, wherever it may be, we want to make sure that we understand the pulse of the fan, to make that experience the best it can be, to guide appropriately, to hold the hands of our fans as we get closer and closer. So to be able to utilize this and synthesize the data in an instant is just something that we find to be a dream. So I think at this point we should look under the hood and have Gavin and Hunter show what they built.

Sounds good. Awesome. Gavin, I think is going to start off first going through it.

Yep. So yes, thanks Jen and Doug. When Doug and Jen came to us with the use case of needing to increase the frequency of their data sets which were previously running just daily and they need to increase them while maintaining efficiency for these live events. There were two things that came to mind. One being, let's just increase the data sets to run every fifteen minutes because of course that's the best case scenario for that, but then you're sacrificing on the credit efficiency. And so then we thought of building this pro code app which we could set event times and trigger data sets from here. But to do so, to do this intelligently, we needed to bring together four different Domo layers, one being the pro code side of course, which then pairs with app DB within Domo and then in coordination with workflows and code engine were able to build a whole experience that they can set when these live events are. And then data sets can trigger only when they wanted to at a frequent rate so that of course they can get real time data to make real time actionable insights.

So just kind of show you the apps process and how we did it there. Of course able to set their events. They can import their calendar for when their events are that they want to have frequent updates for. Then this is a simple one with just three data sets that they need, you know, to have these fan insights coming in for. But you can imagine if you have, you know, five to ten different data sets in your pipeline that you need to be coordinated and reduced data latency. But after configuring the app with their events and their data sets, they can add whatever data set or data flow IDs they need. Those are then saved in an app DB, which is then the source of how this all operates.

So then we jump into workflows, make sure my tab here, and this is really simple one, that its sole purpose is to allow the whole brains of the operation, which is the code engine function to continue running and to trigger it so that it can at all times kind of detect and look at the AppDB to say, hey, are we in a live event or NBA Finals Game five or for example, an NBA draft? Are we doing that right now? Because if so, then it's able to every fifteen minutes trigger updates to those data sets, wait for the longest one to finish running so they're both done. Then it triggers the next step and so on and so forth. So just to wrap it all up and let me just show you this code engine function as well. Not to get too deep into it, but then once we wrap all this up and we jump into API and how like what is the point of all this? The whole point is to get all of these actionable insights with real time data to Doug and Jen and team so that they can then get coordinated with their engineering and product teams to solve these fixes. So then we have the API and alerting system, which I'll then pass off to Hunter to take us under the hood of that.

Awesome. And just for sake of time, we're going to run through a quick example just showing what's happening under the hood with Appy AI. So like Jen talked about earlier, NBA Finals, lots of questions around what was happening during it. I'm going to queue off one of these questions and we're actually just going to look at what's happening under the hood of this AI chat and for the performance that they need with their executives or product team looking for quick answers the next day about what went wrong, how to take those lessons and point them towards the next event. We want to scope the context down of their Zendesk data set, like Doug mentioned, as much as we can. We want to build in the tools so that it's getting the answers as quickly as Jen and Doug need them.

So what's actually happening here behind the scenes is it's going through these rounds of querying the data set based off of our input, and it's trying to scope that context down as quickly as possible within the data set with a few different tools. One to go through the data set, one to pull actual examples of transcripts that people are submitting in. And it's all correlated to these columns within the data set to find what are the problems that fans were experiencing during the NBA Finals between this day and this day and filter it down by product and specific error type. And so now it's probably got one more round to pull this all together. And it's going to give Jen and Doug a very in depth breakdown of all the issues sorted by volume. And we could get into specific transcripts and everything as to what went wrong. But yeah, just call it there, if I may.

This is the butter that I was talking about before working with Domo. Just one proof of concept has turned into a tool we cannot live without because it comes back so elegantly written that we're able to provide so much insight to our leadership in ways we could. It would take weeks to do before. It's just incredible to see.

And I'd also like to add that we've enabled a feature where we can take the output and we could decide, OK, what kind of briefing would we like to give our internal stakeholders? Would we want something more targeted to executive leadership or do we want language and tone that is more relevant to our product and engineering team? So we've enabled ourselves to have that option and to be able to share the data, share the insights based on who our audience is and what type of message we want to deliver, all thanks to Domo.

‍