Reimagining the future of work with Agentic AI

As the data economy enters its next wave, enterprises are increasingly turning to intelligent automation and agentic AI to overcome long-standing challenges in legacy transformation, data unification, and decision intelligence. In this exclusive conversation with Express Computer, Gautam Singh, Head – Analytics, Data and AI at WNS Analytics, shares how the company is redefining enterprise AI with a human-first approach. He explains how they are deploying generative and agentic AI to modernise data ecosystems, accelerate M&A integrations, and move organisations beyond cost-efficiency into sustained value creation. Singh also discusses the rise of AI+HI (Artificial Intelligence + Human Ingenuity) as a next-gen BPM model, enabling domain-driven, hyper-specialised solutions that balance automation with ethical governance. From CRM cleansing in healthcare to personalised research intelligence tools, Singh reveals how they are not just enabling digital transformation, but reimagining the future of work itself.

As many enterprises struggle with outdated digital infrastructure, what are the biggest barriers you observe in data modernisation, especially for legacy-heavy sectors? How is WNS addressing these challenges through AI—particularly generative and agentic AI—to help clients leapfrog traditional transformation hurdles?

The biggest barriers extend beyond technical debt. Legacy systems, often 20+ years old, cannot support real-time data and AI workloads. Challenges include low digital readiness, skill gaps, resistance to change, unclear ROI and data security concerns during cloud transitions.

We combine domain excellence with an end-to-end consulting-to-execution approach across data, analytics and AI. Our award-winning AI Utilities Hub accelerates modernisation through reusable, microservices-based components enriched with industry context – enabling hyperspecialised agents for domain-specific solutions.  

We leverage generative AI to summarise unstructured data, accelerate insight generation and automate content-heavy workflows. These capabilities drive organisations from reactive to proactive, data-driven operations. Our AI Utilities Hub also supports Extract-Transform-Load (ETL), data migration and data quality management using generative AI tools. 

A key example is where we helped a healthcare firm modernise CRM data using agentic AI-driven cleansing. Reinforcement learning-enabled AI agents automated manual corrections, freeing up sales teams to focus on customer engagement. This approach improves accuracy while continuously learning for better results. 

With the growing trend of mergers and acquisitions, how can enterprises effectively unify diverse and often conflicting data ecosystems? What role do intelligent data models and AI-driven governance play in turning post-M&A complexity into a strategic advantage?

Mergers and acquisitions (M&A) create a perfect storm of data complexity that goes far beyond technical integration challenges. When organisations merge, they are not just combining databases, they are reconciling fundamentally different ways of defining business reality. For example, one company’s definition of a “qualified lead” might differ entirely from another’s, and these semantic misalignments can derail strategic decision-making long after a merger.

At WNS Analytics, we enable organisations to identify, build and implement a unified data strategy using proprietary frameworks and consulting expertise. Our generative AI-enabled assets support use cases across the data engineering and management value chain. In M&A scenarios, we play a critical role in guiding organisations to develop new governance policies that respect the distinct identities of the merging entities. Our proprietary solutions and toolkits enable AI-driven governance by establishing the right AI strategy and implementing guardrails for the responsible use of AI.

WNS’ AI+HI model is increasingly seen as a benchmark in next-gen BPM. Can you elaborate on how this approach transforms business outcomes—moving from cost-efficiency to value creation?

WNS’ AI+HI model (Artificial Intelligence plus Human Ingenuity) embodies our belief that true transformation comes from balancing machine intelligence with human judgment. Backed by 25+ years of domain excellence across 10 industries, we combine data science and AI capabilities to build solutions that go beyond cost efficiency and drive tangible business value. 

Our human expertise extends beyond process knowledge to include data scientists and engineers, product managers, AI experts, prompt engineers and business consultants. This multidisciplinary model ensures every AI initiative is grounded in real-world business context and delivers measurable outcomes.

For example, we enabled a global insurer to transform its claims recovery process. Our domain experts identified that subrogation opportunities were being missed and recommended deploying proprietary AI models to discover the specifics for further action. The result: a 15-20% uplift in recovery conversions, 97%+ claim detection accuracy and 40% reduction in recovery lifecycle – proving that AI and human ingenuity together unlock enduring business value.

Agentic AI is gaining traction for its ability to act autonomously in dynamic environments. How is WNS deploying agentic AI across business processes, and what tangible improvements have you seen in areas like decision intelligence and service personalisation?

Agentic AI enables autonomous action across complex business processes. At WNS Analytics, we apply our proprietary GAIN framework for agentic AI to help organisations assess and implement the right degree of autonomy based on business goals, compliance requirements and operational complexity. With expertise across 10 industries – many with stringent regulatory needs – we build contextual roadmaps to embed agentic AI responsibly. 

Our approach uses hyperspecialised agents to act, adapt and evolve in enterprise environments, while ensuring autonomy is governed by business context and human oversight. This calibrated model delivers faster, more accurate decision intelligence, enabling personalised services and proactive risk mitigation without compromising control or transparency.

For example, our award-winning SKENSE-based research assistant accelerates insight generation from complex workflows. Built on a secure, multi-agent architecture, it breaks down business queries, extracts and validates data from 20+ sources and delivers generative AI-powered reports with 99% data accuracy. In a recent deployment, it cut report turnaround time by 85% and reduced costs by 92%, transforming manual research processes into intelligent, scalable insight delivery.