• 50th #ebaytechtalk: Soothing the Pain When Handling Table Views on iOS

    Office Club Berlin Prenzlauer Berg

    50th #ebaytechtalk: Soothing the Pain When Handling Table Views on iOS Abstract ----------- When Steve Jobs was fired from Apple in the eighties, he went on to found NeXT and build a new operating system from the ground up. In the early nineties, NeXT laid the foundations of what would later become OS X / macOS and subsequently iOS. Along for the ride came a data source–driven table view that has some very cool properties, but can be painful for developers to handle. In this talk, I will analyze the root cause of these problems and present strategies for dealing with them in a robust way. I will even give you a turn-key solution suited for a more challenging subset of uses. The topics discussed will also apply to collection views. Speaker bio ---------------- Roddi Deecke is the Engineering Leader for the Android and iOS team at mobile.de. After spending the last 20 years writing client-side software on Windows, macOS and iOS, he is heavily invested in software that runs well on the user’s hardware and provides a smooth interface.

  • 49th #ebaytechtalk : Vespa — the open-source big data serving engine

    Office Club Berlin Prenzlauer Berg

    Offline and stream processing of big data sets can be done with tools such as Hadoop, Spark and Storm, but what if you need to process big data at the time a user is making a request? This talk introduces Vespa - the open-source big data serving engine. Vespa allows you to search, organize and evaluate machine-learned models from tools such as TensorFlow over large, evolving data sets with latencies in the tens of milliseconds. Vespa is behind the recommendation, ad targeting and search engines at Yahoo, where it handles billions of daily queries over billions of documents, and was recently open-sourced at http://vespa.ai. Jon Bratseth is a distinguished architect at Oath (formerly Yahoo), and the architect and one of the main contributors to Vespa, the open big data serving engine. Jon has 20 years experience as an architect and programmer on large distributed systems.

  • 48th #ebaytechtalk: Breaking Down Your React App

    Office Club Berlin Prenzlauer Berg

    48th #ebaytechtalk: Breaking Down Your React App “Jedem Anfang wohnt ein Zauber inne” – there’s magic in every new beginning. Building a new web app on a green field is every developer’s dream. A dream that can turn into a nightmare as the codebase grows, matures and finally becomes unmanageable. React is based on the concept of components. The idea is to compose your application of independent building blocks that you can easily reuse or replace with better ones. If you embrace this idea, you can keep your platform tidy and maintainable. If you don’t – well, you can write React spaghetti code just as easily as you can write jQuery spaghetti code. Patrick shares the story of his own green field project, the relaunch of the online community MOTOR-TALK. He talks about his team’s constant effort to “componentize all the things”, to find best practices, strategies and patterns to constantly improve the code base. You’ll take away hands-on advice from real-world experience, to help you make your own code a little better every day, too! Bio: Patrick Hund, lead frontend developer at mobile.de Patrick helps React and Node.js conquer the tech stack at eBay in Berlin. He has been a professional software developer since 2000 and joined eBay in 2010. Notable projects include relaunching the homepage of eBay’s car trading platform mobile.de in 2015 and the ongoing rewrite of eBay’s automotive online community MOTOR-TALK.de (since 2017).

    1
  • 47th #ebaytechtalk: Deep Learning for Recommender Systems

    Office Club Berlin Prenzlauer Berg

    47th #ebaytechtalk: Deep Learning for Recommender Systems Abstract ----------- “People that bought X also bought Y”, “Discover Weekly”, “Top 10 for You” — online services personalized by recommendations are all around us. Online giants like Amazon, Netflix and Spotify use Recommender Systems and Machine Learning to personalize content. This development is accompanied by another remarkable trend in learning machines: Deep Learning. Almost daily we hear about the tremendous progress rumored to make humans obsolete in more and more domains: skin cancer detection, No-Limit-Hold'em Heads-Up Poker, or speech recognition (no, not yet another Alpha-Go example). For his master thesis at inovex, Marcel Kurovski studied the application of Deep Learning for Recommender Systems. He developed a system for vehicle recommendation in TensorFlow based on data from a large German online market that extends Google’s Wide and Deep Learning approach. In his talk, Marcel will share his insights in this emerging research area and those of other big players. Speaker bio ---------------- Marcel Kurovski, Big Data Scientist at inovex Marcel specializes in deep learning and its application to recommender systems. He studied Industrial Engineering and Management at the Karlsruhe Institute of Technology (KIT) where he focused on machine learning, simulation and operations research.

    3
  • 46th #ebaytechtalk: Detecting activity patterns w/ Convolutional Neural Networks

    Office Club Berlin Prenzlauer Berg

    46th #ebaytechtalk: Detecting activity patterns in accelerometer data using Convolutional Neural Networks, a talk by Kaitao Yang Abstract ----------- Neural Networks have become a state-of-the-art method in machine learning these days. Possible applications include image recognition (faces, objects) and the interpretation of large quantities of data. One of these applications is recognizing different states of activity (such as walking, jogging, cycling, driving, sitting on the train, and sleeping) from accelerometer data as present in modern fitness-trackers. Convolutional Neural Networks (CNNs) are a subcategory of neural networks. Successful variants include LeNet, AlexNet, VGG, GoogLeNet, ResNet, Xception, and MobileNet. All of these were originally designed to process two-dimensional data such as images. In this talk I will show that, with minor modifications, the VGG model can be used to process accelerometer data, which is one-dimensional. We trained three CNNs using labelled accelerometer data to detect daily activities. All our trained models achieved 98% accuracy. More details of this study include: (1) data pre-processing (streaming, shuffling, augmentation); (3) activation function; (4) loss function; (5) GPU computing; (6) Dutch supercomputer for training a large-scale model; and (7) Google cloud services (storage and virtual machine). Speaker bio ---------------- Kaitao Yang works as a Data Scientist & Deep Learning Expert at eBay in Amsterdam. Before joining eBay, he worked as a Deep Learning Engineer at Jheronimus Academy of Data Science. He obtained his Professional Doctorate in Engineering (PDEng) degree from the Electrical Engineering department of Technische Universiteit Eindhoven, and his Master degree from the Cognitive Science department of Xiamen University, in 2016 and 2012, respectively. He has a mastery of popular Deep Learning models (CNNs, sequential models, generative models), with experience in applying them to process both structured data (such as SQL and CSV) and unstructured data (images, text, biomedical signals, and stock prices). In addition, he is actively offering Deep Learning lectures. For details, see: https://www.linkedin.com/feed/update/urn:li:activity:6285216778280927233.

  • 45th #ebaytechtalk: Humanising digital. Moving from UX to human context.

    Office Club Berlin Prenzlauer Berg

    What is Service Design and how can it help to build and roll-out the new product or service? In the era of eco-systems, when the product can not exist without a service, the end-customer is driven by the experience. Service Design is a discipline, that helps find the right sweet spots that the new solution should tackle, at the same time helping co-create the omni-channel experience around it. By taking holistic approach to customer lifecycle and the real customer needs, Service Design focuses on designing value chains between all the involved stakeholders. The Talk will be built around sharing Service Design methodology for in-depth user research, cross-functional co-creation and setting the right priorities for building and implementing solution. Based on the broad experience within automotive, financial, insurance, entertainment and retail industries, Alexandra will share short-cuts and tricks how to design experience across touchpoints from a customer and organisation perspective. Bio: Alexandra Agafonova, Digital Service Designer at Deloitte After graduating from Köln International School of Design, Alexandra has been working intensively in the field of Service Design in Netherlands, UK and Russia, running contextual ethnographic research and designing change within public and private sector. At Deloitte Alexandra is responsible for developing Service Design capabilities as well as for designing experience strategies in complex end-to-end digital projects.

    3
  • Designing UX w/Award Winner: Apple, Samsung, PayPal, Intuit, eBay! (paid event)

    Named one of “10 UX Influencers to Follow Now” by Speckyboy design magazine, learn how to design great UX with recent Apple UX Design Director and award winner, Willy Lai. He is teaching a 2-day UX design workshop in Berlin, 24-25 October, 2017 at H4 Hotel Berlin Alexanderplatz. The workshop draws extensively from his two decades of leading UX design at top tech companies in Silicon Valley such as Apple, Samsung, PayPal, eBay, and Intuit, as well as his UX design teaching and mentoring experiences, including UC Berkeley and Stanford University. This fun and immersive workshop is ideal for designers and non-designers, from novice to advanced skill levels. Participants will form teams and learn through lectures, hands-on exercises, and class discussions: ● How to design great UX across devices through hands-on team exercises ● All phases of the UX design process and all layers of UX ● The business value and ROI of UX design, including case studies and statistics ● Design thinking, strategy, user-centered design, and industry-proven principles, concepts, and methodologies ● Strategically aligning user desirability, business viability, and technology feasibility ● Successful cross-functional collaboration, including leadership and influencing skills ● UX career coaching, including transitioning and growing a career in UX PLEASE REGISTER VIA REGONLINE (SEE LINK BELOW). This event requires a paid ticket. Those without paid tickets will not be admitted. Ask your employer for training budget. Discounted group rates available for 2 or more. Enter promo code "UXB20" for additional savings of 20% off! https://www.regonline.com/DesigningUXWillyLaiBerlin

  • 44th #ebaytechtalk: Like a DevOps

    Office Club Berlin Prenzlauer Berg

    Recognize, analyze and solve live problems with JDK and open-source APM tools. Best before release. It’s done. Your application is in production. But now the software begins to change through usage and continuous development. In this permanent state of flux, it’s important to track the evolution of the application’s runtime behavior and the cumulative effect of changes over time, and to be able to react quickly when necessary. The talk is about hunting and solving real-world bugs with standard JDK tools and MoSKito, an open-source performance monitoring tool. BIO Leon Rosenberg is a software architect and performance specialist, who has spent the last 15 years working on optimization of highly scalable portals like FriendScout24, Parship, C-Date or Allyouneedfresh. He is also an author and committer on various open-source projects with a focus on distributed computing / microservices and application performance management (APM).

    1
  • 43rd #ebaytechtalk: Enterprise Kubernetes + Clusters with OpenShift

    Office Club Berlin Prenzlauer Berg

    Lutz Lange on Running Enterprise Kubernetes + Clusters with the OpenShift Container Platform by Red Hat This talk provides an overview of the challenges, solutions and architecture decisions you'll need to tackle if you want to run Kubernetes Clusters. We will focus on Red Hat's OpenShift Container Platform solution as an example of how to solve these challenges. If you get the benefit of Dockerization and Linux Containers, but are wondering how to take it to the next level, then this talk is for you. Basic Kubernetes know-how is benetifical but not required to follow this content. Lutz Lange works as Solution Architect at Red Hat. Docker, Kubernetes and OpenShift are his specialties. For the past 1.5 years he has primarily worked with service providers to introduce them to and support them with the infrastructure part of these technologies.

  • 42nd #ebaytechtalk: Patterns for Calculating User Profiles in Real Time -I.Mazor

    Office Club Berlin Prenzlauer Berg

    Igor Mazor on Design Patterns for Calculating User Profiles in Real Time It’s important to know your users’ preferences and behaviour in the E-marketplace world. If you can quickly understand who your users are, you can optimize the user journey on the E-marketplace platform by presenting relevant products to the user, and by improving the relevance of search results. One way to leverage user preferences is to calculate a passive user profile based on the user’s interactions with the E-marketplace platform. At mobile.de, Germany's largest online vehicle marketplace, we include in the user profile information such as likelihood of a user to select different car colors, price distribution, mileage distribution, etc. The real challenge is designing a scalable system that can calculate profiles for different users in real-time and serve those profiles via a REST API to other stakeholders. At mobile.de, we reviewed some of the most popular open-source stream processing solutions to consider possible architecture designs for the problem. In a nutshell updating a user profile in real time is actually a stateful stream processing system in which the state is the user profile, the state key could be the user ID, and the state update operation can be as simple as counter increment or an average/variance update. The main focus of this talk will be: how to store the stream state? Which is one of the most important choices to be made before designing a stateful stream processing system. During this talk I will compare a local storage option and an external (global) storage option, and will try to present the different trade-offs as a result from selecting each of the mentioned options above. Igor is a Senior Data Engineer at mobile.de, Germany's largest online vehicle marketplace. Since joining the company in 2016, he has been pioneering the implementation of scalable micro-services around data products. Igor’s current focus is on designing and building a real-time infrastructure that can support millions of users per day. Before joining mobile.de, Igor has been a BI developer and a data engineer at sparks42 and Rocket Internet, working on scalable DWH infrastructure and data ingestion pipelines. Igor enjoys solving hard data problems, especially problems which require creative solutions, and he gets satisfaction from seeing his solutions working in a scalable manner, providing the desired results, and helping to drive company business goals.

    5