Managing the challenges of machine learning

The benefits of machine learning for businesses are immense. However, some organizations are feeling bogged down by it. Some don’t see enough benefit from machine learning while others are running into issues related to its development and deployment. To manage these adoption-related challenges, businesses need to keep some tricks up their sleeves.

At DXC, we pay particular attention to developing analytics solutions that are scalable, reusable and focused on addressing the right business metric.

Here are some examples:

  • Industrialized Machine Learning. In a project for recommending pricing of spare parts for an automobile OEM, reusability was achieved by keeping data processing tasks outside of modeling and by dynamically referencing variables in data through indexes in place of hardcoded names. This makes these models reusable across markets, with minimal effort.
  • Exploratory Data Analysis (EDA). In a server failure prediction project, EDA revealed that reliable indicators of failure incidents were unavailable. The Analytics Data Labs team at DXC then worked with domain experts to identify these incidents by parsing ticket descriptions to identify spikes in application related incidents.

Read more about How to Mine Big Data like a Pro at Wired.

Rags Raghavendra headshotRags Raghavendra is head of the DXC Analytics Data Labs. He is responsible for building a world class data sciences team with specific focus on enabling clients enhance their Analytics IQ through deployment of advanced analytics solutions. These solutions are based on foundation of AI, machine learning & domain based IP. In his spare time, Rags is passionate about using data to improve his golf scores, personal health & knowledge of world history.

Bharathan Shamasundar headshotBharathan Shamasundar is a Senior Data Scientist for DXC Analytics. He is a hands-on analytics professional with a balanced mix of business and technology expertise. Bharathan has 12 years of experience spanning delivery and consulting in the functional areas of Manufacturing, Direct Marketing, and Finance & Accounting.


Get a jumpstart on analytics

Artificial intelligence in banking: How to be more trustworthy, efficient and helpful

DXC Labs: Using data stories to accelerate machine learning solutions


  1. Mahim Jain says:

    Very nice thoughts. In fact in one of my projects server failure prediction data was not available. However through EDA and association rules we found many useful insights like clusters of servers where incidents are raised.

Speak Your Mind


This site uses Akismet to reduce spam. Learn how your comment data is processed.