Back

Speaker "Chintan Turakhia" Details Back

 

Topic

Spatio-temporal Forecasting at Uber Scale

Abstract

Marketplace is the brain behind Uber. To achieve optimal efficiency, Uber’s two-sided marketplace of riders and drivers requires real-time algorithms that can learn and adapt. Pricing millions of requests and matching riders to drivers in a constantly changing physical world is no small feat. Even more challenging is doing this in both space and time. Learn how we use machine learning at scale to enable Uber’s Marketplace teams to see into the future and learn from the past.

Profile

Chintan is currently an engineering leader within Uber's Marketplace organization focused on building and scaling algorithms for real-time decision systems on the ride-sharing platform. Previously, he spent 12 years at Qualcomm as a wireless communications expert in 3G and 4G systems. He focused on optimizing Tier 1 wireless networks using machine learning and helped established the efforts at Qualcomm to enable learned intelligence on mobile devices and network infrastructure. He is also a UCLA graduate (Go Bruins!) and a cyclist.