MetaScale: Hadoop and NoSQL Platform Services

We met MetaScale last year and have written about them on our CTOvision.com site, and have had the pleasure of interviewing their senior leadership. We believe they are worth tracking very closely because of their proven past performance. When they say they can do something it is because they have done it and that is important to note. They provide services to a wide range of clients, but were born in the world of big data analytics and data solutions for Sears holdings.

MetaScale offers Big Data as a Service (BDaaS) to build, operate and support a custom-designed computing platform and solution stack for running Hadoop and NoSQL technologies within a customer’s data center. They provide a big data appliance and associated managed services for a Hadoop or NoSQL cluster that allows customers to build custom analytics applications.

For a view into MetaScale and how their approach works, see the video at this link and embedded below:

For more see http://www.metascale.com/

Find them on Twitter at: @MetaScaleLLC

Our August 2013 interview with Aashish Chandra of MetaScale is pasted below for your reference:

An Interview with Aashish Chandra of MetaScale: Big Data Center of Excellence for Sears Holdings

aashishMetaScale is a wholly owned subsidiary of Sears Holdings with a very interesting story. They are a Big Data center of excellence with proven past performance. Born of service to the Sears family of firms, their lessons learned and ability to get things done the right way, fast, can now be leveraged by others.

I recently had a chance to interview Aashish Chandra of MetaScale and captured some insights you may find of interest below. Aashish is a Divisional VP and head of Applications Modernization at Sears Holdings, and Head of Legacy Modernization at MetaScale.

As an introduction: One of the innovative solutions they developed was an automated tool for off-loading mainframe batch processing to Hadoop, which has enabled Sears to save millions in mainframe expenses.

From our interview with Aashish Chandra:

CTOvision: Aashish, thank you for the time today. Can you give us some background on Sears Holdings?

Chandra: Sears Holdings is the parent of companies like Sears, Kmart and Lands End. We have a network of more than 2000 retail stores across the U.S.  We also have a very large online presence.

CTOvision: I can tell by that statement you must have plenty of data.

Chandra: Yes there is data across all our holdings and we have long used it for decision-making. But too frequently our legacy systems were traditional “batch mode” mainframes that did not afford us the agility our business needed. This business need has been the driver of our big data activities, including the formation of MetaScale.

CTOvision: How did Sears Holdings move towards a decision to form MetaScale?

Chandra: Project teams across Sears Holdings have long leveraged open source and many had been working with Apache Hadoop, but about two and a half years ago we started focused experimenting and proof of concepts designed to improve our business agility. Business reasons compelled us to move in directions that are more real time and less batch, and this required new approaches.

CTOvision: What are some of these business reasons?

Chandra: At a high level, our IT challenges stemmed from a data architecture that could not support our need to make more rapid decisions and serve customers.

CTOvision: Can you discuss some of the use cases MetaScale has worked with?

Chandra: A key example of a system that really had to be modernized to better serve our customers is our very extensive customer loyalty rewards program. We needed more agility to better serve our customers and that meant moving from batch mentalities to fast analysis approaches.

Modernization for cost savings is also a key use case. By smartly leveraging modern Hadoop Clusters we were able to reduce cost of data storage, which makes this Hadoop-centric big data approach very virtuous.

Agility is a use case applied to multiple segments of our business. By modernizing the technology and moving away from the batch mentalities associated with old mainframes we were able to engineer much more agile business processes in multiple areas. Our tech savvy shoppers have evolved and have new expectations of our retail and online stores and we had to engineer for agility so our tech could meet the needs of our customers. Our agility use cases have helped us provide enhanced digital engagement, better personalized service to consumers, and better, more engaging experiences online and off.

CTOvision: How does MetaScale work?

Chandra: We leverage our deep understanding of the complexities of Apache Hadoop to help companies accelerate their big data efforts. We can repeat for others the successes we have provided across Sears Holdings by taking the guesswork out of building Hadoop environments so companies can expoit their data holdings quickly.

CTOvision: So you are like a center of excellence available commercially?

Chandra: Exactly. We are a Hadoop-centric center of excellence with a professional team experienced in hard to find skills like MapReduce, Pig, Hive and the many related tools in a Hadoop Cluster. We run several large clusters ourselves with over 1 PB effective storage and have the experience required to help others with their data holdings.

CTOvision: What Hadoop distribution do you use?

Chandra: We leverage Cloudera’s CDH. We can work with any distribution of course.

CTOvision: What do you see as the top challenge facing enterprises regarding Hadoop?

Chandra: I believe the biggest challenge is that business and IT leaders do not know how to build use cases to leverage the power of the Hadoop framework. We believe awareness of what others have done is an important step in helping enterprises think through use cases, and we also believe MetaScale can help organizations think through the right use cases for their deployment.

 

For more see: MetaScale

Leave a Reply