For a diversified enterprise, digital transformation is rarely about finding a single technology that works everywhere.
The challenge is more fundamental: how do you create common digital capabilities without forcing fundamentally different businesses into the same operating mould?
At DS Group, that challenge spans more than 20 manufacturing plants, over 15 business verticals, and a technology landscape of 75+ enterprise applications serving more than 5,000 employees and business users.
For Santosh Singh, Senior Vice President – IT, DS Group, the answer lies in a deceptively simple idea: centralise what should be common, while allowing technology to adapt to what makes each business different.
The group’s centralised IT function serves the entire organisation, creating common standards around infrastructure, applications, cybersecurity and data. Yet the business realities remain highly diverse—from products and supply chains to procurement cycles and shelf lives.
“A dairy business has to procure raw material every day,” Singh explains. “In spices, procurement can be seasonal.” The technology backbone may be shared, but the business logic cannot always be standardised.
That tension between centralisation and flexibility is increasingly shaping DS Group’s journey towards becoming a more data-driven and AI-enabled enterprise.
One IT backbone, many business realities
At first glance, a diversified business can appear to be a collection of separate technology problems. But Singh sees significant commonality beneath the surface.
Infrastructure is shared. Cybersecurity can be centralised. Enterprise applications generate data that can increasingly be brought together. Finance, for example, follows common regulatory and reporting principles even when businesses differ.
“The IT function caters to the entire group,” Singh says. Centralisation, he argues, makes it easier to establish processes and standardise capabilities across the organisation.
The complexity emerges at the business edge. A dairy operation may require immediate information for day-to-day decisions, while another business may operate on a different planning cycle. Products have different shelf lives. Procurement patterns vary, as do customer expectations.
Technology, therefore, has to be standardised without becoming rigid.
The same principle applies to customer engagement. As consumers interact through e-commerce, traditional channels and increasingly location-specific digital experiences, DS Group is using technology to strengthen last-mile reach, listen to customer signals and enable more personalised engagement.
The strategic challenge is no longer merely digitising processes. It is connecting these interactions into a system that can help the enterprise respond faster.
The data problem behind the AI opportunity
With more than 75 enterprise applications in operation, DS Group has no shortage of information. The harder question is whether that information is sufficiently connected, clean, and accessible to support intelligent decision-making.
“Since we have so many applications in place, now the data is coming—how do I consume this data effectively?” Singh points out. The answer is a growing focus on building a stronger data lake and AI layer.
But collecting data is only the beginning. Information generated within systems the organisation directly controls may be relatively reliable. The challenge increases as DS Group captures data from second- and third-tier participants in the supply chain and from a diverse retail ecosystem.
The objective, Singh says, is to transform raw information into a “gold standard” through data validation, cleansing and testing before it is used for analysis and decision-making.
The larger goal is not simply better dashboards. It is trust.
If business users do not trust the information, they will not make decisions based on it. And without trusted data, AI becomes another layer of technology without a reliable foundation.
Moving AI from individual productivity to enterprise capability
DS Group’s AI journey reflects a similar progression.
The first phase focused on individual adoption, allowing employees and functions to explore AI tools and understand where they could improve productivity. The next question was more strategic: how can those capabilities be scaled across the enterprise rather than remaining isolated experiments?
“AI is here, it will stay, it will have an impact on us,” Singh says. In his view, the technology has moved beyond its initial hype cycle and is now evolving rapidly towards more practical enterprise use.
The emerging use cases are already visible in core business functions. In sales and marketing, AI and data analytics are helping teams monitor whether digital campaigns are translating into commercial outcomes and make corrections faster. What previously required more time for analysis is moving closer to real-time visibility.
In procurement, the opportunity is equally significant. DS Group can bring together external signals—from weather and crop conditions to commercial pricing—to support decisions on what to buy, when to buy, and at what price. “Sales and procurement are the two areas where you have seen the most impact,” Singh adds.
AI is also changing the IT function itself. According to Singh, coding timelines that earlier stretched into months have, in some cases, been compressed to days or even hours, reducing waiting time for internal users.
The lesson is that AI’s first impact may not always be a dramatic new customer-facing product. Sometimes, it is the quiet acceleration of hundreds of everyday decisions and development tasks.
Adoption before optimisation
One of Singh’s more interesting perspectives concerns AI economics. While enterprises are increasingly debating compute costs, token consumption and ROI, DS Group’s immediate priority is finding the right use cases and driving adoption.
“We have seen the power of AI. It is creating impact,” he says.
That does not mean cost governance is being ignored. Singh describes emerging mechanisms around approved queries, reusing previously generated outputs and establishing budgets at individual, departmental or enterprise levels.
But his sequencing is deliberate: first ensure that people understand how to use the technology effectively; then optimise consumption and cost.
This places people—not models—at the centre of the transformation.
“You may have the best technology, but if your people are not equipped or aligned and are not able to adapt, it doesn’t make sense,” Singh argues.
For DS Group, AI literacy is therefore becoming an organisational capability, not merely an IT training programme.
The road to an AI-ready enterprise
Looking ahead, Singh’s technology agenda is organised around a set of connected priorities.
First, AI-enabled applications are becoming an important consideration in technology evaluations. New tools are increasingly expected to include AI capabilities or at least have them clearly established on their product roadmap.
Second, data privacy and security are moving to the centre of enterprise architecture. As the group expands its use of data, including IoT-generated information, it is reviewing applications and platforms to strengthen alignment with India’s evolving DPDP requirements.
Third, DS Group is working to bring more enterprise data onto its data platform while creating access controls that ensure information is visible only to the people authorised to use it.
Fourth is the cloud. The group began with public-facing applications and is extending its cloud journey as more AI-enabled applications and capabilities are delivered through cloud environments.
Finally—and perhaps most importantly—is continuous learning. “Whatever I have learned today probably may not be relevant after six months,” Singh avers, reflecting the extraordinary pace of AI’s evolution.
For a company operating across dozens of technology platforms, multiple businesses and thousands of users, that may be the defining challenge of the next phase.
DS Group’s transformation will not be determined by whether it adopts a particular AI model or application first.
It will depend on whether it can build something more enduring: a common technology foundation, trusted data, secure access, cloud-scale capabilities—and a workforce capable of continuously learning how to use them.
In that sense, the group’s AI strategy is not about replacing its diverse businesses with one digital template.
It is about giving each of them a common intelligence layer—and allowing that intelligence to become more valuable with every decision.