Otto GmbH is Germany’s largest e-commerce site for home and living, and one of the first retailers to take its mail-order business online more than 25 years ago. Today, customers of the Hamburg-based company search 3 million products from 6,800 brands, and make 55 percent of purchases on smartphones and tablets. Otto’s data scientists use object-centric processing to analyze customer interactions across channels for a personalized shopping experience. Its team turned to Exasol as they began to process increasing volumes of data.
Otto planned to use Exasol as an access layer, taking data from a data lake and warehouse to process reports. But the feeder system data structures retained twice as much data, requiring an overly complex loading path. They needed a solution behind the data warehouse to get data to Tableau quickly.
Midway through the PoC, Otto replaced its data warehouse with Exasol, retaining Hadoop as a raw data store. We quickly learned Otto’s security requirements meant going beyond a new database. Together with their team, we implemented a spark connector as part of a joint hackathon with Otto, Exasol, and partners.
“Exasol’s analytics database proved impressive. Primarily due to an excellent price-performance ratio: low costs coupled with exceptional performance and integration with Hadoop.”Andreas Bonet, Product Owner at Otto
With its new Exasol architecture, Otto teams report:
Data is stored and processed directly in the main memory with Exasol’s in-memory technology and parallel processing. Thousands of Otto users can access the data at the same time without sacrificing speed or performance
Intelligent algorithms monitor and auto-tune database performance. Otto’s teams can integrate a variety of programming languages via user defined functions (UDF)
with Hadoop, Tableau, and other frameworks for data preparation
Otto can scale with Exasol’s architecture flexibility
for Otto teams to access data and coordinate campaigns
can sell products at otto.de