California-based software company MapR, which this year saw its approach to Hadoop adopted by Google for its Compute Engine, is kicking off operations in Europe.
It opened an office in London and Germany this week, placing new employees on the ground to target sales in the area and boost relationships it already has formed with partners such as EMC and Cisco.
MapR was formed about three years ago and one of its co-founders came from Google. Unlike other platforms that build on Hadoop with open source code, MapR charges for its platform with the promise of an enterprise-grade product.
MapR VP of Marketing Jack Norris said the company, which has been in operation for three years, remained in “stealth mode” for two of these as it worked on innovations around its code to gain advantages around the Hadoop-based software built on for its Big Data analytics.
Norris said MapR tried to make Hadoop easier to use, with customizable dashboards, self-healing, snapshots and full automated states for failover.
“This is what is really taking Hadoop to the next level,” Norris said.
Apache Hadoop is an open-source software framework for data-intensive distributed applications which can carry out map/reduce and provide a distributed file system that can store data on compute nodes.
It was first designed to support the distribution for Yahoo’s Nutch search engine project. Today Hadoop is used by Facebook, ebay and Google among others.
More interestingly, in recent times, Norris said MapR’s growth has come from the financial services industry, major retailers, manufacturers and government agencies.
“These companies are all looking at leveraging new data sources, part of this unstructured, which can range from click stream data to social media to sensor data,” Norris said.
“In manufacturing this could be information from industrial equipment to increase uptime. MapR allows for self-healing, which means you can automatically recover and do rolling upgrades. It involves a real-time response to business operations and business situations, it is not about analysing what happened last quarter,” Norris says.
Norris said people will be willing to pay for MapR’s advanced features, which make it more applicable to a broad set of mission-critical applications.
“We also have some services we provide, and a network of partners who can connect with our customers,” Norris said.
“What we do is we take the open-source software of Hadoop and combine it with our own innovations to include the open-source code. Things like HBase, Hive and Zookeeper are part of the MapR distribution. It is basically a big data platform you can support a variety of operations on top of.”
Norris said recently, the company has seen customers building their own applications ad algorithms on top of the MapR platform.