Case Study: Connexity achieves real-time ad insights and slashes data processing from days to hours with Cloudera

Connexity Complements the EDW with Cloudera to Improve Retail Insights

Connexity (formerly Shopzilla) is an online retail pioneer that connects more than 50 million shoppers to over 100 million products across a global portfolio of comparison-shopping and review sites. Rapid growth had overwhelmed its 10-year-old EDW—processing 15,000 retailer feeds and 100 million products per day was taking days, data science and BI were starved for fresh data, and the 500 TB Oracle warehouse was growing by 5 TB daily, creating unacceptable latency.

Connexity implemented a hybrid Big Data platform by augmenting its Oracle EDW with Cloudera Enterprise (CDH), HBase, and a suite of Hadoop tools (Sqoop, Hive, Pig, Impala, Spark), plus a custom Forklift utility and Cloudera Manager for cluster operations. The new environment cut processing from days to hours or minutes, enabled real‑time reporting on 10 billion ad bid requests daily, supports scoring 10 million keywords per day, and delivered faster, more actionable insights for data science, advertising optimization, and monetization.

Connexity

Rony Sawdayi

Vice President of Engineering