Information Visualization for Large-Scale Data Workflows
Michael Conover,Senior Data Scientist, LinkedIn
The ability to instrument and interrogate data as it moves through a processing pipeline is fundamental to effective machine learning at scale. Applied in this capacity, information visualization technologies drive product innovation, shorten iteration cycles, reduce uncertainty, and ultimately improve the performance of predictive models. It can be challenging, however, to understand where in a workflow to employ data visualization, and, once committed to doing so, developing revealing visualizations that suggest clear next steps can be similarly daunting. In this talk we’ll describe the role that information visualization technologies play in the LinkedIn data science ecosystem, and explore best practices for understanding the structure of large-scale data in a production environment. From hypothesis generation and feature development to model evaluation and tooling, visualization is at the heart of LinkedIn’s machine learning workflows, enabling our data scientists to reason and communicate more effectively. Broken down into clear, structured insights based on proven technology and workflow patterns, this talk will help you understand how to apply information visualization to the analytical challenges you encounter every day.
Solving The NoSQL Data Dilemma
Will Wilson, Senior Engineer, FoundationDB
Many applications depend on diverse data structures that need to be stored and queried. Developers face a dilemma: they can shoehorn different kinds of data – such as relational, graph-structured, and semi-structured – into a single store, or they can incur the operational cost of provisioning and running multiple DBMSs and their associated technology stacks. This talk will explore a new approach for overcoming this dilemma: mapping multiple logical data models onto a single storage substrate - a transactional ordered key-value store. This approach provides the flexibility of polyglot persistence without its operational complexity. Databases that effectively support multiple data models can provide better options than more specialized solutions such as graph-based storage formats. In particular, multiple query/traversal patterns can be supported with the same physical representation.
API Zen for App Developers
Steven Willmott, CEO and founder, 3scale
Best practices on building, deploying and maintaining API driven applications (mobile and web). At 3scale, one of the world’s leading cloud-based API Management solution providers, we power over 350 APIs- that’s a lot of apps!- and thus we have a vast experience on what problems API Providers and App Developers Face. So our unique vision makes us a perfect fit to give a workshop like this.
API Design Workflow
Jakub Nesetril, CEO, Apiary
API and UI shares a surprising amount of similarities that are not obvious at first glance. An API is essentially a UI, where the user happens to be a developer. Most APIs these days are first designed on a whiteboard, then fully implemented and then exposed to the first external users. We'll explore together a superior approach. Like UIs, APIs should go through rapid design iterations that are exposed to end-users and modified based on user behavior.
Keeping User Generated Content Flexible with Neo4j
Greg Jordan, Founder, Graph Story
In this session, we'll examine how to use Neo4j to keep connected, user-generated content flexible and fast. We'll first take a look at how Neo4j can help achieve a manageable & dynamic structure for content as well as how it compares to our experience with other persistence alternatives. Next, we'll do quick overview on using PHP, Ruby and .Net + Neo4j as well as cloud service offerings. Finally, we’ll share tips & caveats in migrating from mysql, some interesting queries and next steps in expanding use of Neo4j.
It’s All About Me: Techniques for Driving Truly Personal Information Experiences
YY Lee, COO, FirstRain
Personalization is becoming an assumed part of technology UX. Rapid advances—and increased consumer expectations—create both a need and an opportunity for enterprise software to deliver personally relevant experiences in the context of traditional workflows. To close the gap towards a high-impact experience for everyone, we need techniques to discern individual care-abouts from heterogenous, unstructured, overly sparse or noisy, or self-conflicting data. This session will discuss techniques for building user profiles, leveraging explicit vs. implicit factors and the challenges of ambiguity—and how to discern what a user truly cares about to create a highly adaptive, individualized information experience.
Single Sign-On Authentication in the World of Rich User Data & Analytics
Vivek Sodera, Co-Founder & CEO, Airseed
Technology platforms now exist that enable brands and app developers to easily authenticate their users cross-platform through the web and mobile. That user authentication is coupled with opt-in access to hyper-rich transactional, interest, and behavioral data and analytics about those users. Is it possible for brands and developers to consume the depth and breadth of this rich data, especially for real-time personalization, while being good stewards of their users' privacy?
The Art and Science of Machine Learning: Balancing Man vs. Machine
Kimberly Nevala, Director - Business Strategies, SAS Best Practices, SAS Institute, Inc.
Big data has provided the impetus for machine learning (not a new concept) to transcend traditional boundaries. Today companies are utilizing machine learning to exploit and apply insights from the internet of things and traditional ‘small data’ sources alike. Machine learning provides the ability to systematically identify and react to key business events and triggers. But it does not absolve human resources from participation in decision making. Using case studies and emerging research we’ll explore the applications and boundaries of machine learning. And provide some guidelines for balancing informed human decision making with machine generated insights. The discussion will include examples of machine learning applications gone right, gone too far and key considerations for determining appropriate applications.
Highly Available Machine Learning Systems to Move Money
Taylor Phillips, Machine Learning Lead, Square, Inc.
Square's flagship innovations, such as frictionless signup/underwriting and next-day settlement, are enabled by powerful machine learning behind scenes. These use cases boil down to moving money in real-time with machine learning; this is why availability and correctness are of highest concern. In this talk, we'll discuss Square's machine learning pipeline and how it let’s us move fast without compromising uptime.
The End of the Data Scientist
Bruno Aziza, Chief Marketing Officer, Alpine Data Labs
Reports of the demise of the data scientist have been greatly exaggerated. Or have they. Some believe that data scientists remain the foundation of the predictive enterprise and companies who want to win with data shouldn’t compromise their data science values. They should be aided by tools, technologies and practices that help them scale their knowledge and talent. The key is not to replace them but rather make their work available broadly across the organization via software. Others believe that data scientists are no longer vital in the predictive enterprise. Business users will be able to perform data analysis and the new visualization tools will make it easy. Recently touted as the hottest job around, are the fortunes turning for data scientists?
Data Visualization, and Responsibility for Data Literacy
Mikko Jarvenpaa, COO, Infogram
A message requires a sender and a recipient. Data visualization and data communication can not be separated from data literacy. There are the immediate technical responsibilities for presenting data – ensuring it is not misleading, or is presented in the right context – but there is the wider social responsibility of ensuring that the audience understands data and the choices made in communicating with and visualizing data. This is further complicated by socio-economic skills gaps in using data for reporting and communications. Like reading and writing are inseparable, working with data is key to understanding data. This talk will present some of the challenges, the steps being taken to take on the responsibility of improving data literacy across the world, and some implications of this.
Big Data - Tools for Battling Complexity in the Modern World
Gary Robinson,Program Director, IBM
We live in a world of complex systems, with information coming at us faster and faster. Decision making becomes more difficult as we are overwhelmed by the available choices. Healthcare, government, the environment; wherever you look you will find complexity . We need tools to address the complexity inherent in the modern world and tools that can help us make sense of the huge volumes of information at our disposal. Big Data, the next wave of computer systems, offers the promise of machines that will work alongside us, tirelessly consuming and analyzing massive amounts of data, helping us to find the right answers in high failure rate environments. Hear more about Big Data, Cognitive Computing and the exciting innovations that will make tomorrow's computer systems our trusted advisers.
Theseus' Data: Migrating Production Backends with Zero-Downtime
Vijay Ramesh, Software Engineer, Data Science, change.org
As a company grows and changes, so does its data infrastructure needs. Technologies and data-models that may have made sense at one point in time gradually become ill-suited for handling the problems facing your business today. When that data is mission-critical, though, it's often a daunting task to migrate backends or models - leading many engineers to spend countless hours dealing with workarounds, caching strategies, and other mitigation patterns rather than facing the data model problem head on. As change.org has grown from a small blog site to the world's leading petition platform we have had to deal with this problem in a number of ways. This talk will explore how we have successfully transitioned production-facing data from MongoDB and LevelDB to Cassandra and MySQL through a multi-phase process with a focus on data integrity and verification.
Pushing the Boundaries of Computer Vision with an On-Demand Workforce
Seth Teicher, Head of Content & Business Development, CrowdFlower
CrowdFlower makes it easy to tap into an online workforce of millions to label, clean and enrich data at massive scale. During this workshop, we'll share how data scientists use CrowdFlower to rapidly create ground truth for computer vision and machine learning algorithms. We'll explore how scalable human intelligence empowers companies to go beyond what was previously impossible to achieve in facial recognition, wine bottle identification and cellular imagery analysis. What would you do with an on-demand workforce?
How APIs Are Changing The Way We Handle Our Money
Joseph Polverari, General Manager, Yodlee Interactive
As APIs become increasingly accessible and easy to integrate, the financial industry is seeing power shift from traditional banking institutions to quick and nimble startups that are disrupting the way people interact with finances. No longer do we have to write out checks or bother with bank transfers -- Venmo makes peer-to-peer payments as easy as a text message. No longer do small businesses in need of capital need to spend business hours at a bank branch and risk the embarrassment of being turned down for a loan -- Kabbage provides small business lending online, and in a matter of minutes. This presentation will discuss how this power shift is a result of APIs allowing businesses and individuals of all sizes to access financial data that has traditionally been controlled by large, established banks. It will also discuss what this explosion of innovation in the non-bank sector signals for the future of finance -- and what this future will look like.
|DataWeek + API World 2014 is the largest data + API event in San Francisco for the third year running, where thousands of engineers and executives converge to discuss the role of data + API innovation on business, technology, and society. API World 2014 represents standalone tracks and a dedicated tradeshow area of the DataWeek + API World event!||The theme of DataWeek 2014 is Data Skills - and our mission is simple: expose a larger audience of entrepreneurs and executives to data skills like R and Data Science, Data Visualization, implementing NoSQL or Graph Databases, API Design, API Strategy, and Data-Driven marketing. Most of DataWeek + API World will consist of educational bootcamps and workshops, helping bring you into the data age!|
|Sat 13||Sun 14||Mon 15||Tues 16||Wed 17||Thur 18|
|Data + API Hackathon||Workshop + Bootcamp Day||DataWeek + API World Conference & Expo||DataWeek + API World Conference & Expo||Axway API Workshop|
|Technology Hiring Mixer||DataWeek + API World Reception||Data + API Startup Summit|
|Internet of Things Day|
|DataWeek + API World Reception|
|DataWeek + API World Closing Party|
|Tips and tricks to build a cloud based Big DataViz using D3js|
Amazon Web Services
Bank of the West
BASIS Science, Inc.
CA Department of Insurance
NBC Bay Area
Tesla Motors, Inc.