Case Study: NinthDecimal achieves 10x faster cluster throughput with Squadex's Big Data and Machine Learning platform

NinthDecimal gets a clearer picture of consumers’ real-world behavior with modern hybrid cloud Big Data and Machine Learning platform

NinthDecimal, a location-data-powered marketing platform, needed to modernize its legacy analytics environment for real-time, multi-source data processing at massive scale. Squadex was brought in to improve the company’s hybrid data platform, which was struggling with slow handoffs, frequent Spark job failures, and wasted time re-running jobs across a 50-node AWS and 400-node bare-metal Mesos cluster.

Squadex designed and built a modern hybrid cloud big data and machine learning platform using technologies including Apache Kafka, Apache Spark, Amazon EMR, Amazon S3, Amazon Kinesis, Amazon Lambda, Kubernetes, and Snowflake. The new platform processes 5 billion events per day, supports 10 data products, and delivered a 99% job completion rate, 10x higher cluster throughput, and 50% improved team productivity, translating into roughly $100K/month in savings and $1M/year in ROI gains.

NinthDecimal

Lenny Blyukher

Vice President of Engineering