Fresh meal delivery service HelloFresh is using MapR for analytics.
HelloFresh delivers more than 7.5 million meals a month to customers in Australia, Europe and North America.
The Berlin-based company has expanded significantly since it built its own BI system in 2011.
"What we had just wasn't scalable, especially as our needs became more demanding," said chief technology officer Nuno Simaria.
{loadposition stephen08}On the basis of recommendations by other companies, HelloFresh chose Hadoop in the form of the MapR Platform, along with Tableau, Apache Impala and a number of third-party services hosted on Amazon Web Services.
"With the MapR Platform, we can process tens of millions of rows of data in a few minutes. We get results quickly, gain more valuable business insights and can act on them swiftly."
The new system took just two months to implement, including data migration. For comparison, the original home-cooked system took six months to build.
Processing is speedy too: two weeks' data can be crunched in just 20 minutes. That's a big improvement on the nine hours taken by the old system. Faster processing means more opportunities to explore and experiment, leading to greater agility.
Data gathering has been automated across sources including the corporate database, website, distribution centres, customer care tickets and application usage.
In addition to predefined reports, HelloFresh has been able to give its management teams the option of exploring the data in a self-service environment.