Music Streaming Service

"Pythian’s staff have been instrumental in helping us architect and operate the service. They’re immediately and impressively responsive whenever needed, which isn’t often because most times concerns have been proactively identified and resolved prior to degradation. Their teams have helped us shape concerns without having to manage large teams, and I sleep much better at night knowing our core database and Hadoop systems are in very capable hands."

Overview

A popular streaming music service required a platform for its inventory of over 20 million tracks, to cultivate vast amounts of usage data and merge it with curator lists from DJs and machine-learning algorithms. With multiple open source technologies contributing to the service, the company chose Pythian because of its multi-functional support teams that could tackle any technology. Pythian delivered an end-to-end MapR Hadoop infrastructure enabling the platform to withstand a 20-fold spike in demand during its inaugural launch during the Grammy Awards.

PYTHIAN HELPS TRANSFORM STREAMING MUSIC SERVICE WITH HADOOP SOLUTION

Not satisfied with delivering exceptional audio products, a streaming music provider wanted to offer a new listening and musical discovery experience by incorporating human curation—on a completely new platform.

 

Critical Issues

To support an inventory of over 20 million tracks, the platform needed to be flexible and fast, cultivate vast amounts of usage data—including plays, likes, hates, skips, new user information, and social media input—and merge that information with curator lists from professional DJs and machine-learning algorithms. With multiple open source technologies contributing to the service, the company needed a multi-functional support team that could tackle any technology.
 

We Provided

The platform would be the first of its kind in the industry, so to ensure success, the company partnered with Pythian to develop the data infrastructure. They required a big data solution, and it was up to Pythian to determine the best course of action. Pythian experts sat down with in-house teams and identified the tools that would work best to meet the company’s business requirements and accomplish their platform goals. Pythian implemented the data infrastructure and adjusted implementation plans as requirements shifted. Pulling in experts from its Big Data, MySQL, and Site Reliability Engineering (SRE) teams, Pythian achieved the client’s objectives under the tightest of deadlines, while diving into an unknown environment with many technologies and applying a sophisticated, collaborative approach to system management.
 

Result

Pythian tailored an end-to-end MapR Hadoop infrastructure. The company launched its platform the week before the Grammy Awards, receiving significant media attention. Thanks to an aggressive marketing campaign, the platform was flooded with user activity, but easily handled a 20-fold spike in demand.
 

Technologies

MapR Hadoop Kafka Camus Protobuf Hive MySQL SRE
Share: