- Browser Insecurity, Autonomy Visualization, & Lifecycle of a TC39 Proposal
Agenda: 5:45 PM - Doors Open 6:15 PM - Script Alert 1 by Alex Sexton (Stripe) 6:40 PM - AVS: An Open Standard for Autonomy Visualization in the Web by Xiaoji Chen (Uber) 7:05 PM - Lifecycle of a TC39 Proposal by Jordan Harband (Airbnb) 7:35 PM - Networking Passionate about web development? Join us for an evening of talks and networking at the Destination:Web meetup in San Francisco! Script Alert 1 Gone are the days of `.html(userInput)`, but the web ecosystem of 2019 leaves plenty of vulnerable surface area to worry about. Let’s break some websites together. AVS: An Open Standard for Autonomy Visualization in the Web Autonomy Visualization System (AVS) is Uber's open standard for visualizing robotics data. It provides the foundation to building highly complex web applications that playback, interact with, and manipulate data from autonomous vehicles, accelerating the development of self-driving technology at Uber and beyond. Lifecycle of a TC39 Proposal I'll talk about TC39, the proposal stages, about some specific proposals I've championed/participated in/observed, and give an overview of how the audience can get involved Note: We must collect your first and last name for security reasons prior to the event. You will also need to sign an NDA when entering the building.
***Pls RSVP Here: https://www.meetup.com/Go-Silicon-Valley/events/260174483/ Agenda 6:00PM- Doors Open & Social Hour 6:30PM- Streamlining Go Builds We all love Go because it's fast, but what do you do when it stops being fast? In this talk, we'll look at easy things you can do to speed up your Go Docker builds both on your machine and in your CI. With all of these tricks you'll be able to shave up to 80% off your build time, potentially saving hundreds of hours over the next year...and be the hero of developers everywhere. About the Speaker Dan Garfield, Chief Technnology Evangelist at Codefresh Dan Garfield is a full-stack engineer, Google Developer Expert, and member of the Forbes Technology Council. As a Kubernaut and CI/CD expert, Dan has built tools for advanced deployment methodologies with Kubernetes, Helm, and Istio. His code and talks have been featured at conferences including Kubecon, Dev Week, Google Cloud Summit, SwampUp, Redis Conf and many more. https://twitter.com/todaywasawesome 7:00PM- How Uber "Go"es. Elena Morozova from Uber will tell the story of how Go went from being used by a few enthusiastic developers at Uber to becoming the most popular language for microservices in use there. Come learn where they failed and how that led them to solutions that they think are pretty darn neat. About the Speaker Elena Morozova is a software engineer at the Go Platform Team at Uber where she works on Uber's Go ecosystem and wants to make Go experience more enjoyable at Uber. She is a co-organizer of the Women Who Go San Francisco meetup. She is excited about Go, hiking, and dancing. Note: We must collect your first and last name for security reasons prior to the event. You will also need to sign an NDA when entering the building. ***Pls RSVP Here: https://www.meetup.com/Go-Silicon-Valley/events/260174483/
- Advanced Data Analytics with Spark & XGBoost
Spark and XGBoost play critical roles in the landscape of large-scale data processing and machine/deep learning. During this meetup, we invite experts from these fields to discuss how they’ve leveraged these technologies at Uber, NVIDIA, and Intel. Agenda: 5:30PM - Doors Open 6:00PM - Performance Story of XGBoost-Spark by Nan Zhu (Uber) XGBoost-Spark has been the major form of deployment of XGBoost in production environments. Uber’s Nan Zhu will walk through the performance improvements his team has made to XGBoost-Spark across multiple versions, how they’ve led to optimizations in the Uber tech stack, and, as a contributor to the software, discuss a high-level roadmap of XGBoost 1.0. 6:20PM - Analytics Zoo: Distributed TensorFlow and Keras on Apache Spark by Jiao Wang (Intel) Building a model is fun and exciting, but putting it to production is always a different story. While TensorFlow and Keras focus on model building, a complete DL/ML system always needs a robust infrastructure platform for data ingestion, feature extraction, and pipeline management. Apache Spark is a perfect candidate. This talk will provide an overview of Analytics Zoo, a unified analytics and AI platform for distributed TensorFlow, Keras, and BigDL on Apache Spark. 6:40PM - XGBoost on GPUs by Rong Ou (Nvidia) XGBoost is a popular open source machine learning library that solves many data science problems in a fast and accurate way. In this talk, we discuss how XGBoost applications can be significantly accelerated using CUDA-capable GPUs. 7:00PM - Q&A and Networking Note: We must collect your first and last name for security reasons prior to the event. You will also need to sign an NDA when entering the building.
- MoneyCon 2019
*** RSVP here: https://moneycon2019.splashthat.com/ Please join us for MoneyCon 2019, sponsored by Uber. The domain of Payments, Finance, and more generally FinTech, is a fast-growing industry that reached record global investment of $111.8B in 2018. The underlying technology that powers this incredible growth is also evolving rapidly. Hosted by Uber, MoneyCon 2019 will bring together engineers from important tech companies, who will meet and present the latest innovations in their respective fields. Please join us and watch technical presentations by engineers from Uber, Netflix, Square, Ant Financial, Dropbox, Airbnb, and LinkedIn about the Payments and Finance stacks at their companies. With gross bookings of $50 billion for 2018, a global reach of more than 600 cities across 64 countries on six continents, and a constantly expanding list of businesses, Uber must solve a large amount of financial technical challenges. We’re truly excited to be sharing our latest achievements with you and uniting so many important actors of this field for this special event. Food and beverages will be provided. Agenda 5:30 PM Doors Open Sign in, grab some food, drinks and swag! 6:20 PM Keynote Lee Crawford, Uber 6:30 PM Reliable processing in a streaming payment system Emilee Urbanek and Manas Kelshikar, Uber Uber's payment payment is a set of microservices connected with streams. In this talk, we will present some of the challenges we face in operationalizing this architecture and discuss the solutions we developed to address them to build a more reliable payment system. 6:45 PM Network tokens at Netflix Kasia Trapszo and Remi Duvergey, Netflix Netflix embarked in a scheme token journey a couple of years ago, in this talk we will discuss the benefits, the approach we took and the lessons learned along the way. 7:00 PM Books: Scalable, Flexible, and Immutable Storage of Square's Financials Anthony Bishopric, Square Square processes and disburses hundreds of millions of dollars per day. Learn how Square took applied a novel cryptographic concept and classic double entry accounting principles to make it happen. 7:15 PM Break Raffle Time! 7:30 PM Scaling up Cashless: Technical Risk and Quality Engineering Ziying Zheng, AliBaba / Ant Financial This talk will give a high level view of how Ant Financial insures confidence and quality in processing payments at scale. 7:45 PM Evolution of revenue optimization at Dropbox Kirill Sapchuk and Evgeny Skarbovsky, Dropbox We are going to talk about how revenue optimization (fighting involuntary churn) has evolved at Dropbox over last few years from some simpler ideas and experiments around retries up to ML based approach that we’re working on today. 8:00 PM Payment Transaction Routing at LinkedIn Tim Tan, LinkedIn A look into how LinkedIn approaches payment transaction routing taking into consideration local payments, fallbacks, and experimentation. 8:15 PM Controlling Our Own Destiny - Payments as a Service(s) at Airbnb Sophie Behr & Michel Weksler, Airbnb 8:30 PM Raffle & Networking Get ready for some cool prizes! This is a great opportunity to network with our speakers and guests. 9:00 PM See you @ Sports Page Continue the conversation at a bar down the street. 1431 Plymouth St. Mountain View, CA. *** RSVP here: https://moneycon2019.splashthat.com/
- H3: Engineering Sub-City Geos for a Hyper-Local Marketplace with Uber
We use Uber's open source hexagonal grid system, H3, to maintain the health of the Uber marketplace. We use sets of hexagons known as clusters as a flexible way to define geos coarser than a hexagon and finer than an entire city. Come learn about the key concepts of H3, how we use hexagon clusters to identify city cores, and about the data platform we built to serve hexagon clustering. Agenda: 5:30PM - Doors Open 6:00PM - Introducing H3 6:20PM - Engineering an H3-based Geospatial Data Platform at Uber 6:40PM - Building City Cores with H3 7:00PM - Q&A and Networking More on the talks: Introducing H3 by Nick Rabinowitz H3 is Uber’s open source hexagonal hierarchical geospatial indexing system. H3 is used to index geospatial data in Uber’s marketplaces, supporting analysis, operations, and machine learning. This introduction will cover the basics of the system, including how the grid is constructed, common grid operations, and an overview of the kinds of use cases H3 enables. Engineering an H3-based Geospatial Data Platform at Uber by Guocheng Xie & Ankit Mehta Uber operates in the physical world, making tons of decisions to set people and cars in motion, both in real-time and offline. H3 has gained its popularity due to its rich support at Uber and has now deeply rooted in Uber's data set. Marketplace Intelligence team is building an efficient and scalable H3-based geospatial data platform. Building City Cores with H3 by Marie-Camille Achard & Camilla Nawaz We often focus our analyses, experiments, and products on the places where our ridesharing business is the densest. We refer to these areas as city cores. In the past, these were often manually drawn. Using H3 hexagon grid and computer vision techniques, we created a data-driven set of core areas. Come learn about the methodology and how we visualize the heart of our cities! Note: We must collect your first and last name for security reasons prior to the event. You will also need to sign an NDA when entering the building.
- Ludwig: a code-free deep learning toolbox
*** Please RSVP here: https://www.eventbrite.com/e/uber-ai-tech-talk-tickets-61281760395 In this talk we introduce Ludwig, a deep learning toolbox that lets people without a machine learning background train prediction models without the need to write code. Ludwig is unique in its ability to help make deep learning easier to understand for non-experts and enable faster model improvement iteration cycles for experienced machine learning developers and researchers alike. By using Ludwig, experts and researchers can simplify the prototyping process and streamline data processing so that they can focus on developing deep learning architectures. Agenda: 5:30pm - 6:00pm: Mix & Food 6:00pm - 6:15pm: Introdution 6:15pm - 7:15pm: Ludwig: a code-free deep learning toolbox, by Piero Molino, Uber 7:15pm - 7:45pm: Q&A and Networking 8:00pm: Close Speakers: Piero Molino: Piero Molino is a Senior Research Scientist at Uber AI. He works on natural language understanding and conversational AI. He is a co-founder of Uber AI. Note: We must collect your first and last name for security reasons prior to the event. You will also need to sign an NDA when entering the building. *** Please RSVP here: https://www.eventbrite.com/e/uber-ai-tech-talk-tickets-61281760395
- Data Science Toolkits - Apache Zeppelin, Uber’s Ludwig & Data Science Workbench
This is the inaugural meetup for Ludwig, Uber’s new Open Source toolbox. Built on top of TensorFlow, Ludwig lets data scientists and other people who might not be experts in machine learning train and test deep learning models without the need to write code. In this meetup, we will show how Ludwig and the Data Science Workbench are being used at Uber. We will also introduce our friends from the Apache Zeppelin project, who will show off its new features that help productionalize machine learning models. If you are a data scientist, don’t miss this event! Agenda 5:30PM - Doors Open 6:00PM- Intro 6:15PM - Ludwig: a code-free deep learning toolbox (Piero Molino - Uber) 6:35PM - From raw data to informed intelligence: democratizing data science at Uber with DSW (Atul Gupte - Uber) 6:55PM - Preview 0.9.0 - ML model serving from Notebook (Moon Soo Lee - Apache Zeppelin) 7:15PM - Q&A and Networking Note: We must collect your first and last name for security reasons prior to the event. You will also need to sign an NDA when entering the building.
- Marketplace Dynamics: Principles & Technologies That Power Matching and Pricing
The Marketplace Dynamics team at Uber builds the products responsible for the real-time execution and optimization of the Uber Marketplace. We invite you to an evening of technical talks featuring the technologies behind the matching and pricing systems that power Uber’s ridesharing business and make our services more scalable and reliable. We will walk through how we create our matching algorithms to optimize driver ETA and better understand user demand, as well as discuss how our pricing technology is designed to meet the different needs of both riders and driver-partners alike. Learn more at https://marketplace.uber.com/ Agenda 6:00 PM - Doors Open 6:30 PM - Introduction 6:40 PM - How we match Riders and Driver-Partners at Uber (Xinxi Chen) 6:50 PM - Scaling Uber’s real-time optimization with Flink (Xingzhong Xu) 7:00 PM - Rider Incentives at Uber (Qi Wang, Mo Xu) 7:10 PM - Processing Geospatial Incentives at Scale (Sevag Frankian) 7:20 PM - Improve Driver Earnings by Providing Real-Time Geospatial Positioning Advice (Di Zhu) 7:30 PM - Q&A and Networking More on the talks: How we match Riders and Driver-Partners at Uber (Xinxi Chen) When a rider presses the “request” button, Uber actively decides which driver-partner to connect them with for the most seamless and efficient Marketplace match. This talk introduces Uber’s matching system, shows how Uber’s dispatching algorithms connect riders and driver-partners, and discusses how we enable better uberX and uberPOOL experiences for users worldwide. Scaling Uber’s real-time optimization with Flink (Xingzhong Xu) Dynamic pricing leverages real-time data to balance the Uber’s marketplace. This tech talk will introduce the technical challenges of optimizing a two-sided marketplace in real time and share how we leverage Apache Flink to scale up pricing optimization at the global level. Enhancing the Rider Experience with Incentives at Uber (Qi Wang, Mo Xu) In this presentation, we will discuss how our Marketplace team leverages the machine learning models and system architecture to enhance the rider experience with incentives. Learn how we are using machine learning to power rider promotions. Processing Geospatial Insights at Scale (Sevag Frankian) Learn how we scale complex event stream processing systems for high accuracy, low latency, and error correction, providing the most reliable information to power our geospatial intelligence systems. Providing Real-Time Geospatial Positioning Insights at Scale (Di Zhu) In between trips, Uber recommends areas to driver-partners where they will encounter more pick-up opportunities, enabling them to increase their earnings and offer lower wait times for riders. This talk covers the engineering framework that powers these positioning suggestions. Note: We must collect your first and last name for security reasons prior to the event. You will also need to sign an NDA when entering the building.
- IT Eng at Uber: Client Platform Endpoint Management
How do you deploy new management tools remotely to ~30,000 devices? How do you streamline the granting of access to employees while simultaneously securing their devices? How do you scale and replicate the setup of employee devices for a hyper-growth environment? This meet up highlights the engineering and silent framework you don’t often hear about beyond the help desk. Client Platform Engineering (CPE) focuses on the automation of deployment, management, and security of laptops and other mobile devices in Uber’s production and corporate environments. CPE also works cross-functionally with many other internal teams, delivering requirements for audio/video, global builds, infrastructure, networking, and security. Join our CPE team as we dive into one of our most recent accomplishments: enrolling all the macOS devices into our Mobile Device Management (MDM) platform for streamlined compliance, provisioning, and manageability. Learn how our IT Engineering team uses a combination of open source, vendor platforms, and in-house tooling as part of our strategy. Agenda: 6:00PM - Doors Open 6:30PM - Introduction to Uber’s IT Eng Org and the Client Platform Engineering team by Luis Madrigal, Engineering Manager 6:45PM - Overview of Mac/Windows Mobile Device Management Deployment by Danielle DiBella and Erik Gomez, CPE Engineers 7:15PM - A dive into newly contributed Open Source Tools for Mac Management at scale by Nate Walck and Erik Gomez, CPE Engineers 7:45PM - Q&A and Networking Note: We must collect your first and last name for security reasons prior to the event. You will also need to sign an NDA when entering the building.