Loading…
Uber Open is a one-day summit for developers and community leaders advancing open source collaboration and innovation at scale. Featuring over 15 talks and workshops delivered by the technologists building and using some of Uber’s most popular open source projects, Uber Open will showcase how these technologies are driving the future of data visualization, artificial intelligence, site reliability, and more. RSVP and more information here:  Uber Open 2018.


Sign up or log in to bookmark your favorites and sync them to your phone or calendar.

Thursday, November 15
 

9:00am PST

Registration & Breakfast
Doors will open at 8:45am for check in. Join us for a light breakfast before the summit starts. 

Thursday November 15, 2018 9:00am - 10:00am PST
All-Golden Gate Bridge Room

10:00am PST

Opening Keynote with Jim Zemlin
Join us for a keynote from Jim Zemlin, Executive Director from the Linux Foundation

Speakers
avatar for Jim Zemlin

Jim Zemlin

Linux Foundation


Thursday November 15, 2018 10:00am - 11:00am PST
All-Golden Gate Bridge Room

11:00am PST

Break
Thursday November 15, 2018 11:00am - 11:15am PST
All-Golden Gate Bridge Room

11:15am PST

Robust Universal Web Applications with Fusion.js
In this talk, we will go over building a simple application with Fusion.js and what facilities are available to ensure quality and robustness while maintaining high velocity. We will compare building out a few features with traditional web frameworks against Fusion.js and show how we leverage dependency injection, static type checking, and a plugin-based architecture to move fast and [not] break things.


Thursday November 15, 2018 11:15am - 11:45am PST
Bridge Room

11:15am PST

Pyro - Deep Universal Probabilistic Programming
Pyro is a deep probabilistic programming language built on PyTorch, a GPU-accelerated deep learning framework. Developed at Uber AI Labs by Noah Goodman and team, Pyro is used as a platform for research in modern Bayesian machine learning, where deep neural networks can be used both in models and in inference. To scale to large datasets and high-dimensional models, Pyro uses stochastic variational inference algorithms and probability distributions built on top of PyTorch. The Pyro team works closely with the PyTorch team and many open source collaborators to create a rich, stable toolset for probabilistic machine learning research.

Speakers
avatar for Fritz Obermeyer

Fritz Obermeyer

Engineering Lead, Pyro, Uber


Thursday November 15, 2018 11:15am - 12:00pm PST
Gate Room

11:45am PST

Using Service Workers for Performance
Service Workers are best known for allowing web applications to work offline, but they can also make your online pages load a lot faster. We will present an overview of Service Workers and demonstrate why they're your friend if you care about performance. We’ll also introduce the Service Worker plugin for Fusion.js.

Speakers

Thursday November 15, 2018 11:45am - 12:05pm PST
Bridge Room

12:00pm PST

Horovod: Distributed Deep Learning in 5 Lines of Python
Horovod is a distributed training framework for TensorFlow, Keras, and PyTorch. In this talk, we’ll show you how to take a single-GPU training script written in your favorite deep learning framework, and scale it to efficiently running on hundreds of GPUs with just a few lines of code.

Speakers
avatar for Travis Addair

Travis Addair

Senior Software Engineer II, Uber Technologies
Travis Addair is a software engineer at Uber working on the Michelangelo machine learning platform. He leads the Horovod project and chairs its Technical Steering Committee within the Linux Foundation.  In the past, he’s worked on scaling machine learning systems at Google and... Read More →


Thursday November 15, 2018 12:00pm - 12:45pm PST
Golden Room

12:05pm PST

Scaling a React Component Library - Building Uber’s new Design Language
Duplication of code and inconsistent design and experience are major issues across web applications. This talk discusses the process of designing and building a high-quality library of reusable, composable, themeable, and highly-customizable React component library to create a consistent design.

Speakers

Thursday November 15, 2018 12:05pm - 12:45pm PST
Bridge Room

12:45pm PST

Lunch
Lunch will be provided for all attendees 

Thursday November 15, 2018 12:45pm - 1:30pm PST
All-Golden Gate Bridge Room

1:30pm PST

Tiling the Earth with Hexagons
In order to build a better version of real-time surge pricing, Uber decided to get rid of its geofences and create an algorithm that uses a global, dynamically aggregating hexagon grid. In this talk, we will tell the story of why Uber created this global hexagon grid and how the grid works.

Speakers
avatar for Cameron Blume

Cameron Blume

Product Manager on H3, Uber
Talk to me about Surge Pricing!


Thursday November 15, 2018 1:30pm - 2:00pm PST
Bridge Room

1:30pm PST

Urban Computing with Advanced Visualization
How can data visualization be used to understand urban phenomena or even predict the future of cities? In this talk, we will cover how powerful open source software developed in-house can help us find answers to questions that today impact quality of life for everyone.

Speakers

Thursday November 15, 2018 1:30pm - 2:15pm PST
Gate Room

1:30pm PST

Petastorm: Training and Evaluation of Deep Learning Models Directly from Datasets in Apache Parquet Format
Petastorm is an open source data access library that enables single machine or distributed training and evaluation of deep learning models directly from datasets in Apache Parquet format. Petastorm supports popular Python-based machine learning frameworks such as Tensorflow, Pytorch, and PySpark. In this talk, we present Petastorm features and show how Petastorm shortens the model development cycle at Uber ATG.

Speakers

Thursday November 15, 2018 1:30pm - 2:15pm PST
Golden Room

2:00pm PST

Hierarchical Hexagons in Depth
Geographical grid systems often require different resolutions to address use cases at different scales, and it can be valuable to have a hierarchical relationship between cells at one resolution and those at the next. H3 has a hierarchical system where parent cells approximately contain their finer resolution children. This talk will describe how this system is constructed, what use cases are and are not approximate, and gives examples of use cases enabled by a hierarchical grid system.

Speakers
avatar for Isaac Brodsky

Isaac Brodsky

Senior Software Engineer, Uber
Isaac Brodsky is a Senior Software Engineer at Uber, with experience in big data and real time analytics. He is the lead for the H3 open source project.


Thursday November 15, 2018 2:00pm - 2:20pm PST
Bridge Room

2:15pm PST

Large-scale Geospatial Visualization with Kepler.gl
TBD

Speakers
avatar for Shan He

Shan He

Sr. Data Visualization Engineer, Uber
Shan is a senior data visualization engineer at Uber. She is the founding member of Uber's data visualization team and creator of kepler.gl. At Uber, Shan builds data tools and platforms to make business intelligence easy to access, creates exploratory data visualizations to faci... Read More →


Thursday November 15, 2018 2:15pm - 3:00pm PST
Gate Room

2:15pm PST

Cadence: The Only Workflow Platform You'll Ever Need
Do you find yourself using queues, databases, and cron services to implement the back-end logic of your business? Does your company employ multiple legacy workflow automation solutions that do not scale? In this talk, we introduce Cadence, Uber’s open source solution that hides most of the complexity of building fault-tolerant, long-running applications. It is also so generic that it can be used to replace almost any existing workflow engine backend while keeping its workflow definition language.

Speakers

Thursday November 15, 2018 2:15pm - 3:00pm PST
Golden Room

2:20pm PST

Spatial Intelligence Using Hex
One of the applications of hexagons is to be able to combine different datasets with different geographic shapes and forms. In this talk, we will walk through an example of how to bring U.S. census data to the hexagon level and then visualize the data to gain insights.

Speakers
avatar for Sina Kashuk

Sina Kashuk

Senior Data Scientist, Uber
Sina Kashuk is a Senior Data Scientist at Uber, with experience in remote sensing and machine learning, as well as computer vision, time series and geospatial analysis. He applied data science to resilience efforts for several years at DataKind, NYU & CUNY. He especially enjoys creating... Read More →


Thursday November 15, 2018 2:20pm - 2:40pm PST
Bridge Room

2:40pm PST

Hexagon Convolution for Data Smoothing and Forecasting
Many applications which consume hexagon data need smoothing due to data. One common smoothing approach is based on k-ring. In this talk, we will present a data smoothing approach based on hexagon convolution. We will also discuss how we build real-time forecasting on top of the hexagon convolution.

Speakers

Thursday November 15, 2018 2:40pm - 3:00pm PST
Bridge Room

3:00pm PST

Break
Thursday November 15, 2018 3:00pm - 3:15pm PST
All-Golden Gate Bridge Room

3:15pm PST

Scaling Uber's Big Data Platform
Uber’s mission is to ignite opportunities by setting the world in motion. To fulfill this mission, Uber relies heavily on making data-driven decisions in every product area and we need to store and process an ever-increasing amount of data, in addition to providing faster, more reliable, and more-performant access. This talk will reflect on the challenges faced with scaling Uber’s Big Data Platform to ingest, store, and serve 100+ PB of data with minute level latency while efficiently utilizing our hardware.

Speakers
avatar for Danny Chen

Danny Chen

Engineering Manager, Uber
Danny is the co-creator & co-architect for the open-sourced Marmaray Hadoop ingestion/dispersal framework. Danny currently leads the Hadoop Data Ingestion & Monitoring team focused on building out scalable systems to manage the ingestion of an increasing amount of raw data into Uber's... Read More →
avatar for Abhi Khune

Abhi Khune

Engineering Director, Uber
Abhi leads the Big Data team at Uber. His team is responsible for ingesting, storing, processing and surfacing data and insights for all of Uber’s diverse data analytics and ML requirements. His team utilizes and contributes to a variety of open source systems such as Apache Spark... Read More →
avatar for Ekanth Sethuramalingam

Ekanth Sethuramalingam

Engineer, Uber
Ekanth leads efficiency initiatives inside Uber's data infrastructure teams. Ekanth is passionate about building platforms and systems that make infra very accessible to users. Prior to Uber, Ekanth was at AWS where he built and launched two new services from ground up. He has also... Read More →
avatar for Reza Shiftehfar

Reza Shiftehfar

Engineering Manager, Uber Technologies Inc
Reza Shiftehfar currently leads Uber’s Hadoop Platform team. His team helps build and grow Uber’s reliable and scalable Big Data platform that serves petabytes of data utilizing technologies such as Apache Hadoop, Apache Hive, Apache Kafka, Apache Spark, and Presto. Reza is one... Read More →


Thursday November 15, 2018 3:15pm - 4:00pm PST
Gate Room

3:15pm PST

H3-js: Hexagons in the Browser
Geographic editing and analysis applications are increasingly being built as web apps, rather than desktop software. The h3-js library brings H3’s hexagon grid to the browser, supporting a variety of interactive use cases and geographic displays. This workshop will walk through a set of examples demonstrating how to use H3 for different kinds of geographic interfaces and applications.

Speakers
avatar for Nick Rabinowitz

Nick Rabinowitz

Maintainer, h3-js, Uber


Thursday November 15, 2018 3:15pm - 4:00pm PST
Golden Room

3:15pm PST

Building Your Own Metrics Pipeline
To facilitate the growth of Uber’s global operations, we need to be able to quickly store and access billions of metrics on our back-end systems. As part of our robust and scalable metrics infrastructure, we built M3, an open source metrics platform that has been in use at Uber for several years now. This talk will walk you through what a general metrics pipeline looks like and how M3 components can be used for the various stages of the pipeline without locking you in to a particular vendor. We will also cover some of the unique challenges we face at Uber’s scale and how the learnings from those can be used by other companies.

Speakers

Thursday November 15, 2018 3:15pm - 4:00pm PST
Bridge Room

4:00pm PST

Data Agility - A Journey to Advanced Analytics and Machine Learning at Scale
Uber’s Data Science Workbench enables engineers to make iterative machine learning and deep learning training seamless on single nodes distributed across our Big Data and GPU clusters, and analyze, visualize, and productionize data analytics jobs and ML models. This talk will walk you through the typical workflow of a data analyst at Uber, how they leverage Big Data for ad hoc and experimental analysis, discover datasets, run interactive queries against our petabyte scale data lake, and prepare data for advanced analytics and machine learning.

Speakers
avatar for Zhongting Hu

Zhongting Hu

Software Engineer, Uber
Software Engineer in Uber Presto Team. Managing and optimizing Uber Presto Infrastructure, Security, Presto Proxy Layer.
avatar for Hari Subramanian

Hari Subramanian

Engineering Manager, Uber
Hari Subramanian leads several efforts within Uber data infrastructure including building a highly scalable cross-datacenter data processing layer using Apache Spark and SQL analytics-as-a-Service using Apache Hive. He also works in the intersection of infrastructure, machine learning... Read More →


Thursday November 15, 2018 4:00pm - 4:45pm PST
Gate Room

4:00pm PST

Large-Scale Geospatial Visualization in the Browser with deck.gl
Visualization and interaction with large-scale geospatial datasets is crucial to understanding the pulsing lives and transportation in our cities. deck.gl is a WebGL-powered data visualization framework and the foundation to many real-time, high-performance web applications that power Uber’s business and beyond. This hands-on workshop will teach you about deck.gl’s core concepts and capabilities via real-world examples.

Speakers

Thursday November 15, 2018 4:00pm - 4:45pm PST
Golden Room

4:00pm PST

4:45pm PST

Break
Thursday November 15, 2018 4:45pm - 5:00pm PST
All-Golden Gate Bridge Room

5:00pm PST

Keynote with Zoubin Ghahramani
Join us for a closing keynote with Zoubin Ghahramani from Uber AI Labs

Speakers


Thursday November 15, 2018 5:00pm - 5:50pm PST
All-Golden Gate Bridge Room

6:00pm PST

Happy Hour
Join us for Happy Hour at Natoma Cabana with food and drinks. Meet the speakers and network with people in the community!

Location: 
Natoma Cabana
90 Natoma St, San Francisco, CA 94105, USA

You will need a Uber Open Summit Badge to get in the event. 

Thursday November 15, 2018 6:00pm - 9:00pm PST
Natoma Cabana 90 Natoma St, San Francisco, CA 94105, USA