Artificial Intelligence – Meet the Team

November 23, 2019 / Global

Uber is really one big Machine Learning problem – determining a route, how much to charge, how to pair riders in Pool – and AI Labs helps find solutions to all these problems and more across the many organizations at Uber. Did you know that the Uber AI group is made up of 5 different teams and 96 employees working on supporting Artificial Intelligence Research for Uber globally? Led by Chief Scientist Zoubin Ghahramani, Uber AI is doing exciting work in areas of research including Natural Language Processing, Bayesian Optimization, Neuroevolution, Reinforcement Learning, Deep Learning, and Computer Vision.

Gurnoor Kang, Senior Software Engineer II

Describe your team to someone outside of Uber?

“Sensing and Perception is at the core of removing friction for our customers by utilizing a rich array of sensors and inferring real-world situations from them. As an example, we are able to infer situations like “whether there was a crash during a trip”, “is driver making progress toward the rider”, “is this a fraudulent trip”, “is the rider in a correct vehicle”, and more. Detecting such situations, specifically, safety-related situations are really important to help Uber’s users feel and be safe on our platform.”

“From a technical point of view, It is a great mix of challenges in system and data engineering, signal processing, and machine learning to deliver reliable sensor data and inferences that enable Uber platform to be accurate. The Sensing and Perception organization are composed of cross-functional teams with expertise in distributed systems to signal processing and ML (Machine Learning), working together to deliver sensor platform that integrates with Rideshare, Eats, Freight, and more.”

What attracted you to Uber?

“In Uber’s AI (artificial Intelligence) – Sensing and Perception team, I saw the opportunity to deliver on impactful safety and risk products using the power of sensor data and inferences that can provide insights about the complex physical world, vehicle crash detection is an example.”

What excites you about the organization you support?

“At Uber scale, solving today’s real-world transportation-related situations using world-class technology, situations such as crash detection and making pickups reliable in the dense urban environments.”

What are the biggest challenges you need to solve?

“The most interesting challenge is the need and impact of detecting situations (crashes, fraud, etc) as they happen, which involves reliably processing sensor data at lower latency. Sensor data is not always uniform and consistent as it is collected from a diverse set of hardware devices. This high volume data set (petabyte-scale) is required to be contextualized to Uber entities in a real-time fashion. Under the hood, our tech stack uses Flink, Spark, Kafka, Cassandra, Tensorflow, Horovod, Hive, and more to glue everything together.”

Learn more about our Engineering or Data Science Teams

Apply for a role on the Team