Formula 1® makes the rules accessible to every fan with Agentforce

Accuracy powered by trusted data

The Tech Director runs on a tightly controlled system of context built on Data 360, which ingests the FIA 2026 rulebook, editorial summaries, and an internal glossary from F1’s internal content management system into a single, trusted data foundation for Agentforce.

When a fan asks a question, Agentforce uses Data 360’s built-in retrieval-augmented generation to quickly pull relevant answers from these complex, unstructured sources. Deterministic workflows mean that Agentforce can only answer using approved FIA rules and F1 content rather than generating speculative or unverified responses from the open web. Every response is grounded in verified information and remains 100% technically accurate.

That level of precision is critical because the agent represents the F1 brand. “We protect the brand at all costs, so we took a lot of time and effort to make sure we could trust Agentforce to represent us well with our fans,” said Kemp.

To maintain that trust, Agentforce is intentionally limited to the 2026 rules and regulations and does not access public data or customer information. If a question falls outside that scope, it clearly acknowledges that it cannot provide an answer.

Agentforce uses Atlas Reasoning Engine to interpret the intent behind each question and determine a fan’s level of expertise. F1 also built a custom work graph that defines avid fans compared to new-era fans, helping tailor the depth, tone, and complexity of each response. This allows Agentforce to deliver a personalized experience that feels natural and conversational without relying on personal data.

The future of fan engagement

F1 sees their Tech Director agent as the starting point for a much broader, real-time fan engagement platform.

In the future, they see an opportunity to expand Agentforce’s knowledge base beyond static regulatory content to include live race telemetry from AWS F1 Insights, FIA race control feeds and team radio, F1TV streaming data, and social media signals — which could give fans instant access to what’s happening on and off the track.

“Ultimately, we want to equip Agentforce with super rich data, such as telemetry data, so it can answer fan questions like, ‘When did Lando Norris last change his tires?’ or, ‘What was Carlos Sainz’s last lap time?’” Kemp said.

At the same time, F1 is looking to connect Agentforce to the systems that power their track-side and online fan experiences. This includes Agentforce Commerce for merchandise and digital storefronts, Agentforce Revenue Management for product and pricing, and Microsoft Dynamics 365 for back-office operations.

Integrating these systems of record with Agentforce would allow it to move beyond answering questions. “We're starting to think about how we’ll pull this together so it makes sense from a fan journey perspective,” Kemp said. “For example, if a fan really likes Kimi Antonelli’s jacket, we want to make it easy for them to ask Agentforce and get a link to buy it.”

Momentum from day one

“The nice thing about Agentforce is we put the data in place, built the agent, and two days later, we were getting the responses,” said Kemp. “With help from Salesforce Professional Services, it only took us six weeks to get it ready for the public.”

During its first race weekend, F1’s Tech Director handled more than 6,000 interactions — proving that fans were eager to engage this new way.

Just as important, fans asked more than 1,300 questions outside the agent’s initial scope. Those questions are shaping F1’s Agentforce roadmap, revealing strong demand for broader capabilities like historical insights and real-time race data.

F1 sees their Tech Director agent as the starting point for a much broader, real-time fan engagement platform. By starting small and proving the model in a high-trust environment, F1 created a strong foundation that will enable them to expand Agentforce into more advanced, revenue-driving experiences in the future.