Big Data

Automate big decisions in a big data world

Share this post:

Big data big decisionsNo longer only the domain of science fiction, artificial intelligence (AI) is poised to impact everything from how steel is produced to how banks recommend financial products and even to how farmers grow lettuce. It could even change how people move around cities and do business.

At its core, AI can be defined as a set of technologies that enable computing systems to sense, comprehend and act. Within 10 years, AI and robotics are expected to make a creative disruption impact estimated of between $14 trillion and $33 trillion dollars in cost reductions across manufacturing and healthcare, enabled by the automation of knowledge work, according to a Bank of America Merrill Lynch report.

Rule engine and expert systems have been essential components of symbolic artificial intelligence for decades. How, then, can this AI technology complement machine learning today? How can rules be applied on big data to augment enterprises’ digital capabilities while bringing transparency to decision making?

Business rules provided by IBM Operational Decision Manager (ODM) have proven successful in implementing eligibility, pricing and fraud detection systems. Whether an organization’s systems are running on the cloud or on premises, transactionally or in a batch, its rule engine can run decision logic in few milliseconds and at scale in a cluster. With a governed approach, business users and developers can validate a new version of a policy and inject it dynamically in running systems without stopping or repackaging them.

Traditional business rule applications process records a few megabytes of data at a time. As solutions move to big data, rules may be applied to terabytes of data. To this end, the integration of ODM with Apache Spark and Hadoop MapReduce can help scale business rules solutions to the world of big data, and combine them with machine learning algorithms.

Benefits of a successful big decision program include the following:

  • Ability for business users to author, test, simulate and deploy corporate decision automation and management services
  • Collaboration across business users, developers and operations teams in a governance schema
  • Continuous and fast decision automation and service delivery
  • Agility in the change of automated corporate policies, including emergency fixes to production environments to respond to new situations, for example, fraud detection
  • Scalability of the decision automation enabled by open source cluster technology including Hadoop and Spark
  • Possibility of the convergence of machine learning and rule algorithms in the same cluster and the same data lake
  • Auditability to justify automated decisions

Any organization, regardless of size, can employ a rules engine and utilize big data to achieve high performance levels and automate its policies to deliver repeatable and transparent decisions also reducing costs. The world of big decisions is agile and fast-paced and enables massive simulations and production of AI workloads with potential combinations of Watson services and Data Science Experience.

To learn more, watch the replay of our webinar “Think big: Scale your business rules solutions up to the world of Big Data” webinar that explores how ODM business rules applied to big data can transform your decision automation strategy.

Add Comment
No Comments

Leave a Reply

Your email address will not be published.Required fields are marked *

More Big Data stories

Why data is the key to IBM being “the enterprise cloud”

In a recent interview with CloudTech, John Considine, IBM general manager of cloud infrastructure services, made the IBM view of cloud abundantly clear. “Here’s the trick: for us, we are the enterprise cloud,” he said. Data is a cornerstone of that way of thinking, Considine said: “One of our theories leading into the cloud, for […]

Continue reading

Does your data deserve a private cloud?

Your business and your data are unique. For that reason, your enterprise architecture must also be tailored to fit the exact needs of your business. When data is involved, you want choices, not tradeoffs. More importantly, you want your solutions to build upon and complement one another. For most companies, the variety of data sources […]

Continue reading