Formula 1 x AWS (F1 Insights)
Eight years of F1 data, still sport's most mature AI laboratory.
Formula 1 and AWS launched F1 Insights in 2018, growing to over 20 ML-driven broadcast metrics (built on Amazon SageMaker) that process 1.1 million telemetry data points per second per car. In 2022, AWS's HPC infrastructure enabled two-car CFD aerodynamic simulations that contributed to the 2022 car redesign. In early 2026, the partnership moved into agentic AI: the Data Accelerator, built on Amazon Bedrock AgentCore, cut new data-source onboarding from up to 8 weeks to roughly 40 minutes. F1's global broadcast audience grew from 500 million fans in 2017 to 826.5 million in 2024. Status: continuous production since 2018, expanding scope through 2026.
Por qué importa
This is the longest continuous production history in the archive — eight years of deployment, not a single announcement. It raises a different question than most recent cases: not whether the technology works, but what happens when an AI vendor becomes embedded deeply enough to co-design the car itself, since AWS's CFD simulations contributed directly to the 2022-2026 car specification. The 2026 shift to agentic AI for internal data operations also signals where the partnership is heading next: from fan-facing insight generation to running F1's own back-office data infrastructure.
Stack tecnológico