Beyond Big Data Analytics: Lessons Learned from Google and Microsoft
I will draw from my experiences at Google and Microsoft to give some examples of how those companies use big data analytics to deliver value internally and to their customers. From Microsoft, I’ll explain the cycle of language modeling analysis and some of the major bottlenecks encountered there. From Google, I’ll touch on the AdPlanner system and some of its challenges as well as how dremel helped the company get a better hold of its data. The goal is to balance the technology side with the end-user experience of such systems and show what value was derived from them at the business level. While there is no single formula, these examples will suggest strategies for deploying big data analytics in your organization in support of your business goals.
Biography of Speaker
Theo Vassilakis is the CEO and founder of a venture-backed stealth-mode startup in the big data analytics space. He spent the last nearly 8 years at Google, where he was most recently a Principal engineer leading an organization of 75 software engineers in data warehousing, visualization, and analysis. His teams were responsible for several published systems such as dremel (the backend for BigQuery) and tenzing, but also for critical internal data generation processes such as revenue/billing reporting for sales, finance, and other business functions. Theo also worked on search and advertising at Google including large-scale machine learning for personalized search ads, early versions of the Doubleclick AdPlanner for display ads, logs data features for Google Webmaster Tools, and search UI like page previews and early prototypes of results-as-you-type. Prior to Google, Theo was an engineer at Microsoft where he developed language and acoustic models for speech recognition, and at Microsoft Research where he built data cleaning and mining features for Microsoft SQL Server. Theo holds a Ph.D. in Mathematics from Brown University and a B.Sc. in Mathematics from Stanford University.