NASSCOM Community

A CIOs Perspective on How Cognitive Computing Can Rev Up the Enterprise

Blog Post created by NASSCOM Community on Jul 8, 2016

At the 4th edition of the NASSCOM Big Data & Analytics Summit (23rd – 24th June) we had Sri Donthi, the CIO of PepsiCo address on the opportunities of Cognitive Computing.

Why this sudden interest in wanting to gather CIOs inputs on Cognitive Computing? Simply because, CIOs are now in a very strong position to unlock growth opportunities, brought about by data explosion which is being harnessed by enterprises, to leverage greater value in a consumer-centric world. Big Data continues to enhance Machine Learning, and provide deep industry insights which are practically limitless. This data, is largely unstructured though.

Just consider the humongous amount of data which is being generated by all the industries put together. How can this be used positively to negotiate a future which masquerades as VUCA? Admittedly, the outcome and full impact is unknown right now, but it boggles the mind when we think of limitless possibilities because of data explosion. Some sources argue that approximately upto 5 quintillions of data is being generated every day. What is most interesting, is that ninety per cent of this volume has been in the last two years only.   

Deeper Understanding of the Consumer

The consumer today has a very high consciousness of value, augmented by the fact that comparisons with rival offerings are easily possible. Sensors, capture data all the time. How can we bring all of this together, to gain a deeper understanding of the consumer? For instance, the volume of data generated from the agricultural sector ranks second only to manufacturing. Imagine the sheer value that can be unlocked!  

A study was conducted on corporates and consumers simultaneously. The aim was to capture, if corporates really understood the consumers’ needs. The outcome was stark. While 81% of corporates proclaimed that they understood the consumer, whereas only 37% from the other side (consumers) seemed to agree. While we understand that there are bound to be difference in perceptions, but not something as wide as this. The gap definitely has to be reduced. Cognitive Computing perhaps, may provide us with the answer of how it can be done. It can seek out unseen patterns which can be decoded for deeper insights about consumer buying behaviour. If this can be achieved, then most certainly there will be a significant positive impact on competitive advantage.

The four stepped approach:

  • Engage through better collaborations
  • Discover new ideas, through Big Data & Analytics
  • Decode, based on evidence backed recommendations
  • Automate – enhance speed, process, quality and integrity of deliverables 

Addressing Mammoth-sized Challenges

28% of the global population is dependent on agriculture. However, this sector alone uses 70% of the water resources. If we had better sensors and many more of them in number, perhaps we could avert disasters due to extreme climatic conditions. There are 8 billion humans on this planet. Can we use all the data that’s being generated, to address the challenge of food scarcity?

Sales Executives at Pepsico have access to numerous reports. If totalled, they are actually repositories of a huge amount of data. Pepsico leverages all of that data to assist Smart Sales Executives with appropriate action plans, which ultimately impacts top line, positively.

The threat of cybersecurity is unprecedented. In 2014 alone, close to 1 billion records were stolen. Some estimates have put the figure even higher. Is there a pattern to these thefts? If it be so, can the next disaster be averted? Can we use Big Data to “out-plan” the hackers?

The Future of Customer Engagement

  • Gather information about consumers
  • Engage with them through multiple touch points
  • Provide a consistent brand experience
  • Offer products to meet consumer preferences

Navigating the Hype Around Cognitive Computing

  • The purpose is to seek answers to critical business problems
  • Develop a roadmap
  • Develop models
  • Train
  • Deploy, explore & evolve
  • Be cognizant of security threats at all times
  • Seek long-term opportunities over short-term ones
  • Execute where you are able to gather mountainous data

 

Outcomes