A Better Match for Drivers and Riders: Reinforcement Learning at Lyft
This addresses the efficiency of matching in ridesharing platforms, benefiting drivers, riders, and the company, and is a documented first implementation of real-time learning in this domain.
Lyft tackled the driver-rider matching problem by implementing a novel online reinforcement learning algorithm that estimates drivers' future earnings in real time, resulting in millions of additional riders served annually and over $30 million per year in incremental revenue.
To better match drivers to riders in our ridesharing application, we revised Lyft's core matching algorithm. We use a novel online reinforcement learning approach that estimates the future earnings of drivers in real time and use this information to find more efficient matches. This change was the first documented implementation of a ridesharing matching algorithm that can learn and improve in real time. We evaluated the new approach during weeks of switchback experimentation in most Lyft markets, and estimated how it benefited drivers, riders, and the platform. In particular, it enabled our drivers to serve millions of additional riders each year, leading to more than $30 million per year in incremental revenue. Lyft rolled out the algorithm globally in 2021.