Biography

Moacyr Passador, Director of Solutions Engineers for the South Region in MicroStrategy, has more than 10 years of Analytics experience. He started his journey in Brazil then he moved to Portugal, Texas, San Francisco and Atlanta. During his time in the West Coast Moa worked with companies like Facebook, Apple, Cisco, Oath Yahoo!, Yelp, Gilead and Wells Fargo with their most interesting Analytics / Big Data projects.

Breakout Session Title: Case Study from Very Large SF / Bay Company

This use case is about understanding your data, technical limitations, and the needs of your internal and external customers in order to develop a comprehensive solution to address the identified challenges of the business. Unstructured data, in nature is challenging; not only in presentation but in how it is managed, often relying upon multiple systems in order to do the job of a traditional DW / Analytic stack. Coupled with how unstructured data scales, elasticity and performance are factors when attempting to provide near time results to your customers.

Who – VERY Large SF / Bay Area Tech Company: That sells advertisements based on clicks and impressions.

Business Challenge – the inability to provide Todays data (clicks and impressions) was impacting advertisement sales when sales reps were meeting with customers.

Technical Challenge – Customer was attempting to run Microstrategy reports on unstructured and structured data based on HP Vertica. Processing time was taking 12 hours due to the size and type of data.

Solution – Migration from Vertica to Presto on Hadoop, w/ Kerberos security and MicroStrategy. Data is accessed and and queried from the source systems using the Presto Query engine allowing for a faster, more stable infrastructure for both structured and unstructured data, cutting reporting times from 12 to 8 hours.

Managing Unstructured Data – Built by Facebook, Presto is an open-souce query engine designed for users looking for fast interactive query capabilities across multiple data sources, (structured, semi and unstructured). Single queries can access multiple data sources from the source systems. MicroStrategy provides the User interface, query and visualization capabilities that would leverage Presto against Hadoop. MicroStrategy’s in memory cubes, provide the performance necessary to manage queries across petabytes of data and against multiple data sources while the Symantic/MetaData layer provides the necessary governance to provide data accuracy and security for the thousands of queries being run daily.