How Chief Risk Officers Are Managing Risk with Hadoop

Contributed by

3 min read

I often get asked, “Who in an organization buys Hadoop?” While there are buyers with many different titles and functions that spearhead the adoption of Hadoop, I’d like to single out one fast growing buyer: the Chief Risk Officer.

Organizations are appointing Chief Risk Officers (CROs) to better manage potential risks that could impede their strategy. These risks can generally be categorized as strategic, reputational, operational, financial, or compliance-related. In the financial services industry, specific risks include property and casualty exposures, foreign currency exchange risks, commodity price fluctuations, and reputational and credit risks.

An emerging weapon for managing these risks is Hadoop. By combining data from different sources and areas of an organization, Hadoop makes it possible to measure and better manage risk. Hadoop “tears down the walls of risk management.” There are many examples of seemingly small exposures on an individual basis that when looked at in aggregate, represent dangerous levels of exposure.

We see companies from a wide range of verticals that are embracing Hadoop. These companies have increasingly large volumes of data (web portal clickstreams, financial records, retailer loyalty program data, social media streams, user- and product-clustering and segmentation) and increasingly narrow timeframes. Many use cases share a common pattern of having multiple weak indicators that in aggregate will indicate upcoming operational problems. Examples of use cases include:

  • Credit and merchant fraud prediction for financial services organizations
  • Cable operators who want to better predict outages and network issues by analyzing set-top boxes
  • Healthcare organization professionals who want to better understand patient outcomes

Given this increased focus on managing risk, the last thing Chief Risk Officers want to do is to introduce risk to the Big Data platform itself. Exposure to downtime, data loss, and outages are unacceptable for this buyer group and a major reason why they are turning to the MapR Distribution for Hadoop for risk management.

This blog post was published August 02, 2013.