The NFL modernizes game-day operations with Microsoft Copilot Studio and Power Platform

There’s more to a National Football League (NFL) game than meets the eye.

Every NFL game is a live, high‑stakes production. Beyond what fans see on the field, thousands of operational tasks must run flawlessly—facility readiness, audio‑visual systems, security, and broadcast support. When something goes wrong, whether it’s a stadium access incident, a malfunctioning microphone, or missing equipment on the sidelines, seconds matter.

A key part of managing operations on the field is recording each incident and developing reports. Previously, this was done during each game using an Excel spreadsheet. The process was time-consuming and, with multiple staff inputting notes, data was often inconsistent, which impacted reporting accuracy, and the ability to learn from past events.

To create a more consistent, data‑driven approach to stadium operations, the NFL worked with its technology partner, Microsoft, to build a unified, AI‑powered reporting solution using Copilot Studio and Power Platform. The Game Ops Dashboard solution provides a single source of truth for game‑day incidents across venues, while enabling efficient post-game reporting and long‑term operational learning.

AI-driven categorization and reporting

At the heart of the solution is a Power Apps application designed specifically for the NFL Game Operations teams. From a phone or tablet, staff can quickly select a specific game and log operational notes or incidents as they occur.

Instead of filling out long forms, staff simply describe the incident—for example, “Mic not working on the sideline” or “Endzone camera feed briefly unavailable during the first quarter.” From there, AI takes over and automatically classifies the incident using a model in AI Builder in Power Automate

As a staff member enters a short description and timestamp for an incident, the model predicts the appropriate category (e.g., broadcasting, uniform policy, game presentation, or stadium facilities), sub‑category, and priority level, along with a confidence score.

These predictions are grounded in domain‑specific knowledge about stadium operations, including historical data of previous incidents which are stored in Dataverse. Using Power Automate, the system also pulls official NFL game schedules and metadata via league APIs. This ensures that every incident is associated with the correct game, teams, venue, and kickoff time.

One of the most powerful aspects of the solution is the AI model’s ability to learn over time. The model recognizes patterns in how different types of incidents are categorized with each incident reported. “As the season went on, we found ourselves making fewer and fewer manual corrections. It was clear that the model was quickly learning and moving to near-perfect accuracy,” says Quentin Autry, Game Operations Associate at the NFL.

The tool streamlines post-weekend reporting by consolidating all incidents after each slate of games into a single standardized view in the app. Built-in filtering allows reports to be tailored to surface incidents by category (e.g. broadcasting incidents) and customize for NFL leadership, or individual teams. Reports can be automatically exported to Excel.

The Game Ops Dashboard app also features an embedded AI-driven agent built in Copilot Studio. The agent provides staff with a fast, convenient way to query data stored in Dataverse. For example, staff can quickly pull up the latest notes for a specific game or week and filter notes by category—all using natural language queries. 

From real-time response to long-term insight

Beyond game‑day operations, the NFL will now have access to a rich historical dataset. Operations leaders can analyze trends such as: which types of incidents occur most frequently; how long different categories take to resolve; which equipment or vendors are associated with repeated failures.

These insights support smarter planning, preventive maintenance, and even vendor negotiations. For example, data can be used to identify recurring equipment incidents and drive conversations with suppliers about reliability, upgrades, or contract terms.

What was once ephemeral, scattered information has become a durable operational asset for the league. “With Copilot Studio and Power Platform, we were able to build a simple, low-code solution, supercharge it with a self-learning AI model, and quickly transform the way we record and manage game day operations,” says Autry. “The solution is saving us valuable time during every game and enabling us to make better operational decisions throughout the year.”

Greg Horrocks, Senior Coordinator, Game Operations, provide this additional feedback, “The platform has helped save the Game Ops team between 8-12 hours a week in data entry and report generation. Additionally, we look forward to the unrecognized efficiencies the platform will provide us as a year-by-year database. This will allow the Game Ops team to be more informed in making policy changes each offseason.”

The NFLs Game Ops solution has also yielded another interesting benefit for an influx of new staff. Rather than spend time up front learning all the categories and sub-categories for incidents, new staff can rely on the app to do the categorization and then learn from the resulting reports. As Autry says, “The app is making it easier for our new recruits to learn the business and make meaningful contributions, earlier.”

As the NFL continues to evolve the platform, future opportunities include deeper predictive analytics and broader application across other league‑managed events. There are also plans to explore expanding use of the agent for operational queries. For example: How many Week 1 games have bad weather delays in the past 10 seasons? What were the most common incidents reported since Week 5 of this season?

What began as a way to streamline incident reporting has become a foundation for smarter, data‑driven stadium operations—helping ensure that, when millions of fans tune in every weekend, the focus stays where it belongs: on the game.