CIO Insider

CIOInsider India Magazine


AI Driven IA - What The Crystal Ball Beholds

Dhaval Pandya, CIO & CDO, JSW Paints Private Limited

A part of the core leadership team of JSW Paints, Dhaval is involved in all the strategic initiatives and discussions across the functions.

Gone are the days when terminologies such as Artificial Intelligence, Neural networks, Machine Learning etc. all were restricted to closed conclaves of Hi-tech labs. These technologies have hit the tipping point where their commercial applications have passed the test and increasingly enterprises are investing into newer avenues of Intelligent Automation driven by Artificial Intelligence. Here’s an attempt to encapsulate the top 3 areas where increasing use cases are expected to emerge in the times to come.

Intelligent Hyper-personalization
It is long acknowledged that 'Customer is the king' and there has been a relentless endeavor by enterprises to know the king better so that they can offer the king just what s/he wants. Right from 'Mera wala pink' to ' two spoons sugar', companies are collating and churning every blink, every smirk, and every like (and dislike) that you make as a part of our everyday lives. The times to come will bring a new breed of platforms to the fore that will crawl and collect every bite of data that maps customer behavior, provides tools built on AI based algorithms that intelligently massage this ocean of data (largely unstructured from myriad sources) and provide business relevant meaningful insights. These insights will then form the foundation stone for companies to devise hyper-personal solutions (not products not services, but solutions) and present them as recommendations to their king.

Their performance will be closely monitored by the very platform which will form the raw material to further sharpen the recommendations. I would call this intelligent platform 'Recommendation-As-A-Service'. Pockets would be dynamically created within this ocean of data basis the primary domain of operation & KPIs to be driven. The levers driving this intelligent platform would be ecosystem driven business models fueled by CAMS (Cloud-Analytics- Mobile-Social) that will allow data capture and sharing smoother than ever. End objective being sought – the king gets his exactwala pink and gives you a 5-STAR review every time.

Intelligent Supply chains
This is an age-old saga to balance supply with demand to make your product/ service available on time every time in the shape and form that it is envisaged to be delivered. With time and advancement, the variables that drive (and influence) demand have moved from being domain specific to non-domain specific to behavior oriented. The move would now be towards intelligent demand creation that focuses on predictions in behavior, changes in 'seemingly unrelated factors' & the environmental conditions hence (not today or past). The PEST (Political-Economic-Social-Technological) data from comparable scenarios would come together to predict the future trends and their due impact on demand.

This would provide companies a peek into the future and gear up their supply chains to be able to service this future intelligent demand at cost. The differentiation would then come from the data models and the algorithms employed and the POC (Percentage of confidence) pegged to the output. Organizations would need to focus on all the aspects of this Intelligent Supply Chain viz 'Intelligence, Supply & Chain'. The complexities would be enormous since a small tweak in an upstream factor will result in multifold changes in business decisions not just for the company but for the entire chain aka the ripple will translate into a wave. The companies that would be able to surf this wave well would sail through.

Cybersecurity has been a hot topic for a while now and the new normal has only accentuated the need for the same

Intelligent security
Cybersecurity has been a hot topic for a while now and the new normal has only accentuated the need for the same. As organizations are moving more and more towards digital models, connected machines and ecosystem driven models, they are also exposing themselves to prying eyes that are waiting for their window of opportunity to make their move. The attacks are also becoming more and more intelligent that are able to exploit more than one type of security flaws and morphing themselves with available resources. Security of the future will be intelligent security that is able to look beyond the traditional signature & DPI based identification mechanisms.

It will need to be prescriptive (not just predictive) and self-healing. Are you thinking SkyNet or Matrix grade...? Well not yet. However, three aspects are most likely to come to the fore. One, it will have an enhanced scope of coverage considering the business models of the future. Two, it will be platform based that can safeguard everything from IT to OT to Endpoints. And three, it will be able to draw inferences from myriad sources real time i.e., the scope of reference libraries & recognition capabilities will no longer be local or point based – it will be global & distributed. Capabilities such as cloud computing & distributed processing would provide the neural network to make this brain function efficiently and impart the ever-increasing intelligence.

That said, a word on the genealogy of this intelligent automation – the way these algorithms are being written will also undergo a fundamental change. It will move to a multi-tier architecture where there is 'AI for AI' and the resources to be deployed would also change dynamically basis the problem being addressed. This will also help this technology break the fundamental constraint of data requirements from needing more & more data to working with optimum data - 'relevant & fit for use'

This is what I see as I gaze at the crystal ball, how it unfolds only time will tell...!!!

Current Issue
Ace Micromatic : Pioneering Excellence in Comprehensive Manufacturing Solutions