Uber is expanding its infrastructure and artificial intelligence (AI) capabilities on Amazon Web Services (AWS). Uber is using AWS Graviton instances to support more of its Trip Serving Zones, the real-time infrastructure behind every ride and delivery, and has started pilot training some AI models on Trainium—enabling faster rider and delivery matching, global demand handling, and smarter, more personalised experiences for millions of daily users.
Every time you open Uber and request a ride or delivery, a series of split-second decisions happens behind the scenes. Which driver is closest? What’s the fastest route? How long will it actually take? Getting those answers right instantly—for millions of people at once—requires the right infrastructure for Uber to deliver these capabilities at scale during rush hour and major events.
How Graviton helps power millions of trips in real time
Uber’s Trip Serving Zones are part of the system that makes sure every ride and delivery runs smoothly, which requires making millions of predictions and processing location data in milliseconds.
Now, Uber is expanding its use of AWS compute, storage, and networking to help power real-time operations for Trip Serving Zones. By running more of these workloads on AWS Graviton, Uber can reduce energy consumption while scaling rapidly during demand spikes, both reducing latency and optimising costs. Graviton’s high performance enables some of the real-time calculations that help match riders with drivers faster—without compromising reliability, availability, or security.
“Uber operates at a scale where milliseconds matter,” said Kamran Zargahi, vice president of engineering at Uber. “Moving more Trip Serving workloads to AWS gives us the flexibility to match riders and drivers faster and handle delivery demand spikes without disruption.”
Improving Uber rides at scale with AWS Trainium chips
Uber has also begun experimenting with AWS Trainium to train some of the AI models that help power its apps. These models analyse data from billions of rides and deliveries to determine which driver or courier to send, calculate arrival times, and recommend the best delivery options to the customer. Training AI at this scale requires enormous computing power—Trainium provides an efficient, cost-effective way to do it. As the models learn from more trips, Uber delivers faster matches, more accurate arrival time estimates, and more personalised recommendations to customers worldwide so they can get where they are going faster and receive their deliveries sooner.
“By starting to pilot some of our AI models on Trainium, we’re building a technology foundation that will make every Uber experience smarter—so we can keep our focus where it belongs: on the people who use Uber every day,” Zargahi said.
“Uber is one of the most demanding real-time applications in the world, and we’re proud to be an important part of the infrastructure powering their global operations,” said Rich Geraffo, vice president and managing director of North America at AWS. “We’re helping Uber deliver the reliability hundreds of millions of people count on today—and the AI-powered experiences that will define ride-sharing and on-demand delivery tomorrow.”
Discussion about this post