In boardrooms across industries, the AI conversation has matured from “Can we?” to “How fast can we? How safely can we?”
What separates leaders from laggards now is not ambition, but execution: the ability to convert data, platforms, and security into daily operational advantage. In this exclusive interview, Santhosh TG, Chief Digital Officer of Switch Mobility, shares a clear, forward-looking view of the organisation’s shift from digital foundations to intelligent enterprises, and why Agentic AI signals a decisive next chapter in enterprise transformation.
Some edited excerpts:
How would you characterize the organisation’s digital maturity today?
We’ve reached a critical inflection point. Our initial focus was building robust digital foundations: enterprise-grade platforms, unified data architecture, and strong cybersecurity protocols. This wasn’t just technology deployment. It was about establishing the operational readiness that separates successful AI implementations from stalled pilots. Industry data shows that 97% of organisations struggle to demonstrate business value from AI initiatives, primarily due to inadequate data readiness and governance frameworks.
That’s why we invested lot of time building these foundations before scaling AI. Today, we’ve transitioned from digital enablement to intelligent operations. The key milestone is moving Agentic AI from proof-of-concept to production deployment across critical business workflows. Digital maturity for us isn’t measured by technology sophistication. It’s measured by execution consistency, adaptation speed, and effectiveness in translating digital capabilities into measurable business outcomes.
What are some of the most critical digital initiatives you are currently driving?
We’re embedding intelligence into core operational workflows from factory floor to enterprise functions like finance and supply chain management. This isn’t about isolated point solutions or department-specific tools. It’s about building integrated platforms where data flows seamlessly across the organisation, enabling faster decision-making and reducing operational exceptions. The strategic shift is in our systems architecture.
We’re deliberately designing platforms that connect and communicate rather than creating isolated silos. When information moves without friction across functions and departments, three things happen: decision velocity accelerates significantly, operational exceptions decrease measurably, and organisational learning compounds over time. In competitive markets where response time determines market position, this architectural approach has direct P&L impact. The goal is reducing the time from insight to action across every business process. That’s where digital transformation creates tangible, measurable business value.
Many organisations struggle to move beyond AI pilots. How are you approaching AI differently?
The primary reason AI pilots fail isn’t technical. It’s operational discipline and organisational readiness. Recent surveys show that 92% of CDOs are concerned about accelerating AI adoption without addressing underlying data and governance challenges. We’ve seen this pattern repeatedly: organisations build impressive demos that never scale because they lack the foundational capabilities required for production deployment.
Our approach treats AI as a long-term strategic capability, not a series of disconnected experiments. We’ve established clear governance structures, secured sustained investment commitment, and defined explicit business metrics for every AI initiative we undertake. Before building any model, we ask critical questions: Where will this reduce cycle time? Where will it eliminate manual touchpoints? Where will it improve accuracy or strengthen compliance? The success isn’t about deploying the most sophisticated AI technology. It’s about deploying AI precisely where it creates measurable business impact.
Could you give us an insight into any AI-based initiative at Switch?
We’ve deployed SMartIE, our Agentic AI-powered digital operator, to improve enterprise-wide productivity. One of its highest-impact applications is in invoice processing. The challenge includes high transaction volumes, frequent exceptions, and multiple data dependencies across POs, GRNs, contracts, and approval hierarchies. Vendors submit data in inconsistent formats.
Information is often incomplete. Traditional automation handles the predictable 60 to 70% of transactions efficiently, but struggles with the remaining 30 to 40% that require contextual judgment. With Agentic AI, we’ve fundamentally shifted the approach. Instead of programming step-by-step workflows, we’ve given the system outcome-level responsibility. SMartIE interprets data, validates against multiple systems, resolves mismatches autonomously, and processes invoices end-to-end. It escalates to human operators only when genuine ambiguity exists. The business impact includes significantly reduced processing time, fewer manual interventions, and improved accuracy rates.
Invoice processing is something many organisations have automated in the past. How does Switch stand out with Agentic AI? How is it different from traditional RPA or workflow automation?
RPA executes predefined rules with high speed and accuracy, but it’s brittle. When exceptions occur, traditional automation breaks down. The fundamental limitation is that RPA follows instructions but doesn’t understand context. Agentic AI operates differently. It interprets context, makes informed decisions, learns from outcomes, and adapts to new patterns. When it encounters an unexpected scenario, it evaluates alternatives, determines appropriate actions, and escalates only when human judgment is genuinely required.
What makes our implementation distinctive is production-scale reliability. SMartIE runs in live production workflows, handling high transaction volumes with minimal human intervention while maintaining accuracy standards. Critically, it owns the outcome. Traditional automation executes steps. Agentic AI ensures results are achieved. It flags risks early, adapts when needed, and escalates strategically. This outcome ownership creates organisational trust and enables scalable adoption.
Finally, what is your vision of an “intelligent enterprise” in practical terms?
An intelligent enterprise learns at the pace of operations, not just during quarterly reviews. It’s an organization where people spend less time on repetitive execution and more time on strategic value creation: improving quality, developing innovations, and solving complex problems. The human role remains central in setting strategic direction, defining policies, and making judgment calls that require contextual understanding.
AI takes over repetitive, high-volume work that can be codified and scaled. The real ROI of AI isn’t cost reduction alone. It’s redirecting human cognitive capacity from execution to innovation. Practically, it means building organizations where systems, people, and processes work in tight coordination so that problems are identified and resolved in near-real-time.
Digital transformation was the journey. Intelligent execution is the destination, and that’s where sustainable competitive advantage is built.