Vertica at the Aeronaut Brewery: Adventures in Data Architecture

Posted January 30, 2018 by Soniya Shah, Information Developer

white cloud in vault type room representing cloud computing
This blog post was authored by Sarah Lemaire.

Vertica hosted a Meetup at the Aeronaut Brewery in Somerville for customers and prospective customers, including data scientists from Nuance, J.Jill, and True Fit. After some cold beer to warm us up on a cold night, we were lucky enough to hear JB Huang, Head of Data Products at Sports Illustrated Play, talk about his experience managing the data architecture at SI Play and GSN Games, and how Vertica impacted it. JB worked at Vertica as the Technical Manager for OEM Sales in a previous life, and has a wealth of experience working with large data infrastructure systems.

For most of the talk, JB discussed his five years at GSN Games and showed how the data architecture there evolved during that time. The problems he faced included:

• High data latency
• Long reporting times
• Dependencies on other teams for data collection
• Systems that were hard to maintain, scale, and debug
• Difficulties tracking new data

GSN Games was growing fast, so to solve these problems, they integrated Vertica into the existing architecture, which helped in the following areas:

• Scalability
• Unified data warehouse
• Data optimized for analytics
• Better analytics functionality
• Faster reports
• Faster parallel loading
• Easier to archive / reload data

JB explained in detail how, over several years and iterations, he was able to improve and simplify GSN Games’ data architecture for more improvements in all these areas.

At the end of his talk, JB described the data architecture for SI Play and answered several questions from the attendees.

Many thanks to JB for taking the time to share his experiences, to the Vertica engineers for answering attendee questions, to Aeronaut for the great beer, and to Sue LeClaire in Marketing for organizing the Meetup.