Uber uses OpenAI to help people earn smarter and book faster

Every day, millions of people rely on Uber to book rides, order meals, send packages, and earn flexibly. Behind every tap is a complex real-time marketplace shaped by traffic, weather, airport arrivals, local events, and demand. Uber operates at massive scale: 40 million trips per day, 10 million drivers and couriers across 15,000 cities in over 70 countries. Each city has its own operating dynamics, regulations, and rider behavior, creating a system that must adapt continuously at global scale.

Uber has long used machine learning to support its marketplace. And now, with the benefit of large language models and OpenAI frontier models, Uber can reason across complex signals more quickly, deliver fast conversational responses, and power voice experiences inside the app.

The collaboration between Uber and OpenAI is helping Uber build AI-powered products that simplify earning opportunities for drivers and couriers and reduce friction for riders. And using OpenAI’s models, Uber can ship streamlined products and experiences faster than ever.

Uber’s engineering teams built a multi-agent architecture that routes each user request to the most appropriate specialized system. For example, earnings questions can be handled differently than onboarding questions, and marketplace guidance requires different reasoning than transactional actions.

This architecture enables Uber to route each task to the model best suited to its specific operational needs, ensuring that each query is handled with the appropriate focus on what matters most.

That has proven especially valuable for new drivers. Uber found that using AI to summarize and easily communicate Uber’s real-world data can accelerate ramp-up by helping drivers learn workflows and marketplace dynamics much faster than through trial and error alone.

Improving support for early lifecycle drivers , helping new drivers position themselves better for more trips

For drivers, flexibility is one of Uber’s biggest strengths. Some drive full-time, others just on weekends, while some drive between classes or shifts. This flexibility also means drivers are constantly evaluating options and asking questions: Where should I position myself right now? Is the airport worth driving to? Should I switch from rides to deliveries during lunch? Why did my earnings look different today?

To help answer those questions, Uber developed Uber Assistant , an AI-powered assistant designed to help drivers throughout their lifecycle on the platform—from onboarding and first trips to day-to-day earnings optimization.

“We want to enable drivers to make better decisions for themselves by providing a summarized view of the marketplace and real-time insights,” says Dharmin Parikh, Director of Product Management at Uber.