NVIDIA's GPU infrastructure and AI platforms will enhance Lyft's enterprise AI capabilities while building towards an autonomous hybrid network

Today at the NVIDIA GTC AI Conference, Lyft (NASDAQ:LYFT) announced it will use NVIDIA AI to enhance the company's machine learning systems across its global operations. The work spans across AI-driven mobility services Lyft continues to invest in, including enterprise AI infrastructure, next-generation mapping systems, and future Level 4 autonomous fleet architectures powered by NVIDIA DRIVE Hyperion.

This press release features multimedia. View the full release here: https://www.businesswire.com/news/home/20260316007778/en/

Lyft & NVIDIA

Lyft will integrate NVIDIA's AI technologies to enhance predictive modeling, advance mapping capabilities, and improve the overall rideshare platform that serves millions of riders and drivers every day.

Powering Lyft's next-generation AI capabilities

The work includes several ​​initiatives where Lyft is applying NVIDIA AI technologies to improve platform performance and operational efficiency:

  • Accelerated computing: By integrating NVIDIA's AI supercomputing capabilities on our cloud infrastructure, Lyft can enhance the performance of compute-intensive workflows that power critical platforms for millions of daily rides. This includes predictive models across critical rideshare systems, as well as large-scale feature engineering pipelines, data transformation processes, and optimization workloads. Incorporating the NVIDIA AI Enterprise suite, which includes NVIDIA Nemotron open models support, NeMo software, RAPIDS Accelerator, and cuOpt, enables Lyft to more efficiently process data in real-time, lowering compute costs and improving rider-driver matching within marketplace operations.
  • Next-gen mapping platform: Advanced vision-language reasoning and multimodal input open new paths for Lyft to improve guidance and routing safety for drivers in real-world road scenarios, and the ability to identify map issues with higher precision. These capabilities could scale the data enrichment process, accelerate core map making, and enhance navigation. As Lyft evolves its mapping infrastructure with agentic AI capabilities, NVIDIA models, including NVIDIA Nemotron family for agentic AI and NVIDIA Cosmos Reasons series for Physical AI, are a natural fit for powering an Agentic AI mapping platform.
  • Supporting Lyft's autonomous vehicle efforts: Lyft continues to pursue autonomous vehicle deployment through multiple technology integrations and collaborations. As part of this work, Lyft plans to leverage NVIDIA DRIVE Hyperion — NVIDIA's end-to-end autonomous vehicle development platform — as a reference architecture for future autonomous fleet systems, integrating high-performance compute, multi-sensor processing, and safety-certified systems.