Ritwik Batabyal-Chief Technology and Innovation Officer at Mastek outlines how enterprises can move beyond AI pilots by reengineering processes, funding innovation through efficiency gains, and building self-optimising, AI-led organisations.
As enterprises move beyond AI experimentation, the focus is shifting toward scalable, outcome-driven transformation. Organisations are under increasing pressure to convert AI investments into tangible business value, while navigating evolving paradigms such as agentic AI, data sovereignty, and intelligent automation at scale.
In this conversation, Ritwik Batabyal, CTO and Innovation Officer at Mastek, explains how the company’s “Led with AI” approach is enabling this shift by reengineering core business processes, embedding intelligence, and validating solutions internally before scaling. He also shares insights on legacy modernisation, monetising AI through efficiency gains, India’s deep-tech opportunity, and the rise of self-optimising enterprises—what he describes as the next phase of enterprise evolution.
CIO&Leaders: Mastek has recently rebranded its approach as “Led with AI.” How are you moving beyond the pilot stage, which many enterprises struggle with, and delivering the 20–30x ROI you claim from AI-led programs?
Ritwik Batabyal: Before discussing outcomes, it is important to explain the process.
We begin by selecting specific internal business processes such as people supply chain, accounts receivable, or accounts payable. These are broken down into journey touchpoints using our ADOPT framework. Each touchpoint is then reimagined using AI or intelligent automation.
We transform the entire process against defined KPIs to achieve measurable outcomes. This transformation is first implemented internally to validate its effectiveness. Once proven, we offer it externally as a software-driven service.
This internal validation allows us to deliver consistent outcomes in external deployments. Therefore, the 20–30x ROI is not a promise but a result of a repeatable engineering discipline. It is essentially a “pilot in action” approach rooted in reengineering business processes.
CIO&Leaders: You have secured significant contracts with the UK Home Office and other public sector entities. In such high-stakes environments, how does your approach ensure that legacy modernisation does not simply migrate inefficiencies to the cloud but actually transforms service delivery?
Ritwik Batabyal: Most organisations approach legacy modernisation as a migration exercise. We take a different approach by reimagining citizen journeys rather than focusing solely on system transformation.
We simplify decision layers, introduce intelligent feedback loops at critical points, and automate compliance-heavy processes. This shifts modernisation from a technical migration exercise to a measurable improvement in speed, transparency, and trust.
Our focus is on outcomes driven by user experience, particularly from a citizen-centric perspective.
CIO&Leaders: The industry is moving from chatbots to agentic AI systems capable of executing multi-step workflows. What architectural challenges do enterprises face in scaling such systems, and how can they manage autonomous decision-making while maintaining trust?
Ritwik Batabyal: Agentic AI is not a standalone technology; it is a combination of multiple interconnected elements. It requires orchestration and the ability to build trust.
Many enterprises struggle because they treat it as a product deployment rather than an integrated system design. Our approach focuses on identifying both data in motion and data at rest, defining implicit decision rules, and embedding governance into the system.
We incorporate observability to monitor agent behavior and decisions, along with human-in-the-loop mechanisms where necessary. These elements collectively ensure orchestration, accountability, and trust.
CIO&Leaders: Cost and revenue remain central concerns. Many enterprises are unable to move beyond experimentation to generate meaningful returns from AI. What is the most effective monetisation strategy for mid-market enterprises?
Ritwik Batabyal: The strategy must be sequential.
The first step is internal cost optimisation. By improving efficiency in business processes, organisations can generate measurable ROI. These gains can then fund innovation.
The next step is reinvesting those savings into AI-driven products or new revenue streams. In essence, innovation should be funded through efficiency gains. This approach ensures sustainability and reduces dependency on external funding.
CIO&Leaders: With increasing regulatory pressure around data localisation and sovereignty, how can enterprises balance centralised data platforms with localised compliance requirements?
Ritwik Batabyal: The future lies in building systems that are globally intelligent but locally compliant.
Data management should leverage centralised intelligence while maintaining strict localised control. This can be achieved through policy-driven data layers, strong encryption, and region-specific governance models.
Such an approach ensures that data remains usable at a global level while adhering to local regulatory requirements.
CIO&Leaders: You have experience across startups and large enterprises. What is the key to building a scalable innovation ecosystem, and how do you select partners?
Ritwik Batabyal: We evaluate partners through two primary lenses.
The first is readiness, which includes:
- Enterprise alignment with the use case
- Depth of domain expertise
- Agility in integration
The second is the operational model, which includes:
- Willingness to invest in early-stage innovation (MVP development)
- Transparency in creating a mutually beneficial commercialisation model
Innovation cannot be purely commercial from day one. Both parties must invest initially to build a sustainable, scalable solution.
CIO&Leaders: India aspires to become a deep-tech hub. What is the one critical gap that needs to be addressed?
Ritwik Batabyal: India’s strengths lie in two areas: service orientation and human capital.
These strengths must be converted into monetisable intellectual property. Historically, the IT industry has focused on labor arbitrage, but the future lies in building IP-driven solutions.
Transforming service capabilities and talent into scalable, commercially viable IP is essential for India to lead in deep tech.
CIO&Leaders: Many industry leaders believe the cloud era is maturing and the focus is shifting toward intelligence. Is there a third transformational phase emerging?
Ritwik Batabyal: The trajectory is moving toward what I would describe as a concept of singularity.
If you look at the evolution, we moved from cloud transformation to broader digital transformation, and now into AI-led transformation. However, beyond these phases, the direction is increasingly toward a self-optimising enterprise.
AI, in this context, is an enabler rather than the end state. The real shift is toward organisations that can continuously optimise themselves. If you reflect on my earlier responses, many of them are centered on optimising internal processes to unlock efficiency, which then enables the creation of new revenue streams.
This continuous internal optimisation is what will define the next phase. Enterprises that can adopt this model effectively, while ensuring that appropriate guardrails, governance, and controls are in place—will emerge as sustainable winners in the long term.
CIO&Leaders: You mentioned the concept of singularity. Could you elaborate on what exactly you mean by that for better clarity?
Ritwik Batabyal: When I refer to singularity in this context, I am essentially talking about a self-sustaining and self-optimising enterprise.
Traditionally, innovation and transformation have relied on a triad involving academia, industry, and customer use cases. There has always been a collaborative architecture where different entities contribute to building and refining solutions.
However, what is now emerging is a shift at the level of business processes themselves. Organisations are beginning to look inward, identifying their internal processes and systematically optimising them in a stepwise manner.
As these internal processes become more intelligent and efficient, the enterprise starts evolving into a more self-reliant entity. This reduces the dependency on external collaborative constructs and enables the organisation to function as a more unified and self-driven system.
That is what I mean by singularity—an enterprise that is capable of driving its own optimisation, learning, and evolution from within, rather than relying predominantly on external ecosystems.