Announcing Uber Open Summit 2018: Collaboration at Scale

Like most technology companies, open source has played a large role in the growth of Uber’s business, and our teams use and contribute to numerous projects. In 2012, we released our first open source project on GitHub. Six years later, Uber now hosts more than 320 open source projects and repositories, with over 1,500 contributors delivering over 70,000 commits.

This global community has collectively created many popular projects, including Jaeger, a distributed tracing system, and Horovod, a distributed training framework for TensorFlow, Keras, and PyTorch. Jaeger, now a Cloud Native Computing Foundation-hosted project, has been on InfoWorld’s awards list for best open source software for cloud computing for the last two years (2017 and 2018). In 2018, Horovod was included in InfoWorld’s awards list for best open source software for machine learning, and is now included in the machine learning suite by most cloud providers.

With global operations in over 600 cities doing over 15 million trips per day, the scale of the technical problems we solve at Uber puts our projects at the boundary of engineering possibilities. For example, RIBs, our cross-platform mobile architecture framework, allows us to scale seamlessly as hundreds of engineers work with the same codebase simultaneously. M3, our large-scale metrics platform for Prometheus, now houses over 6.6 billion time series and aggregates 500 million metrics per second. And our vis.gl suite gives us the tools to map mobility and take the pulse of a city with data visualization.

Understanding the challenges of building technology at scale, we open sourced many projects to foster collaboration with technologists worldwide who are solving similar problems, building better systems at scale one commit at a time. By 2018, we had our mission statement: enabling collaboration through open source. The success of open source is only possible because of this vibrant and giving community.  

As part of this effort to collaborate with our open source community, we are proud to announce Uber Open, a one-day summit this November for open source contributors, users, and community leaders.

The first-ever event will feature over 15 talks and workshops delivered by the technologists building and driving the future of open source at Uber. From data visualization to artificial intelligence and site reliability, our open source work spans multiple domains. Speakers include Nicolas Belmonte, head of the vis.gl data visualization software suite, Fritz Obermeyer, Pyro project lead, Nadiia Dmytrenko, Fusion.js core maintainer, and Felix Cheung, Apache Spark project management committee member (PMC), and many other leaders in Uber’s open source community.

Keynote speakers include Jim Zemlin, executive director of the Linux Foundation, and Zoubin Ghahramani, chief scientist at Uber AI Labs.

This event will also host developers and community leaders from Taiwan as part of the Uber Exchange program announced in August 2018. In collaboration with the Ministry of Science and Technology, Uber Exchange is our flagship knowledge exchange and tech mentoring program for Taiwan to share our knowledge and foster open source collaboration. ¹

We are excited to see you there.

— Brian on behalf of the Uber Open Source Program

To learn more about Uber Open, visit: https://uberopen2018.splashthat.com/

For more information about Uber Open Source, visit: https://opensource.uber.com

¹Uber Exchange is a knowledge exchange and tech mentoring program. In August 2018, Uber announced that Uber Exchange is coming to Taiwan with a special focus on self-driving technology and AI. In collaboration with the Ministry of Science and Technology, Uber will bring high-profile AI experts to Taiwan, sponsor hackathons and open source projects, and welcome Taiwan experts and developers to our flagship facilities.

Editor’s note, 10/4/2018: this article has been updated to reflect Horovod’s inclusion in InfoWorld’s 2018 awards list for best open source software for machine learning.