Three Approaches to Scaling Machine Learning with Uber Seattle Engineering
At an April 2019 meetup on ML and AI at Uber Seattle, members of our engineering team discussed three different approaches to enhancing our ML ecosystem.
Enhancing front-end and back-end experiences
Seattle, USA
Uber is changing the way cities move globally through our transportation and mobility solutions. Located blocks from the iconic Pike Place Market, Uber Seattle furthers our core business objectives across product and platform teams through product, engineering, and operations. The Uber Seattle Engineering teams build technologies that power our driver and rider apps, developer platform, Big Data architecture, and machine learning tools, including Horovod, an open source deep learning framework.
Announcements
Articles
Uber Open Source: Catching Up with Felix Cheung, Data Platform Engineering Manager
Uber Engineering Manager and open source software community member Felix Cheung talks about his work with the Apache Software Foundation, open source at Uber, and XGBoost, a machine learning library for optimized distributed gradient boosting.
NVIDIA: Accelerating Deep Learning with Uber’s Horovod
Horovod, Uber's open source distributed deep learning system, enables NVIDIA to scale model training from one to eight GPUs for their self-driving sensing and perception technologies.
Constructing a Seamless Flow: Meet Uber Engineering’s Developer Experience Team
Get to know Uber's Developer Experience team, a group of writers, educators, and technologists dedicated to setting our engineers up for success.
Meet Uber Engineering Seattle
In this article we take a look inside the Uber Engineering Seattle office, meet a few of the leaders, and learn about the culture and technologies the teams are working on.
Team Member Spotlight
Shawn Burke
Senior Staff Software Engineer
'I think that the stakes and the consequences of what we do here at Uber are much higher than they are at other companies. Like when you have a software problem at another company, that means that you probably lost revenue and maybe you lost some customer trust. When we have a software problem at Uber, that means somebody who is somewhere maybe they don’t really want to be anymore and they want to get home is impacted or somebody is trying to put food on the table. I think those are special things and require you to think about problems in a way that is hard to find in other companies.'
Dana Bright
Design Manager
'At Uber, I love that we’re thinking end-to-end in both the digital and the physical worlds, and looking at the transition between the two. We’re not just thinking about the screen of the Uber app, we’re thinking about everything that physically happens from before and after you’ve even requested an Uber.'
Mon-Chaio Lo
Senior Engineering Manager
'I’m a big collaborator. I truly believe that one of us does not know more than all of us and so, as an engineer, you have to figure out how to utilize your peers and your stakeholders as much as you can. I call it obsessive collaboration because some people only collaborate by need, but a great engineer is building collaboration into the things they are doing every day to ensure they are using all the resources around them.'