EXL started as a traditional outsourcing firm but is now positioning itself as an AI-led data and analytics company. The company is rearchitecting its offerings, talent and business model around AI. According to EXL chairman and CEO Rohit Kapoor, while AI is a huge opportunity with rising enterprise spend, it is only valuable when embedded in business workflows. In this interview, he speaks to Sudhir Chowdhary on how the company is building an end-to-end AI stack. Excerpts:
India has long been the back-office of the world. How is that changing in the AI era?
Many people assume AI will simply automate a large share of back-office work. That is true to some extent, but what often gets overlooked are the new opportunities India’s talent base can unlock. India’s workforce is deeply trained in back-office processes, client operations and industry domains. As AI becomes embedded in operations, that understanding becomes incredibly valuable. India is in a strong position to help global clients embed AI into their workflows, making back-office operations far more intelligent and sophisticated. This will improve global competitiveness, accelerate speed and efficiency, and open up opportunities that would not have existed otherwise.
How has EXL evolved into an AI-driven enterprise?
One of EXL’s strengths has been the ability to anticipate trends and invest ahead of time. In 2006, we acquired a data analytics company called Inductis, long before analytics was widely discussed. We believed that simply running operations would not be enough-companies would need to leverage data insights and convert them into action. Today, analytics accounts for roughly 47% of our revenue in a company that generates around $2 billion annually.
Seven years ago, we started investing heavily in data management because we anticipated that data would become foundational for AI. We built capabilities and acquired companies in this space, and last year launched EXLdata.ai, which uses agentic AI to prepare enterprise data environments for AI readiness. For instance, even before the emergence of ChatGPT in 2022, we had already been investing in machine learning and advanced analytics, positioning us strongly to build on the opportunities created by generative AI.
Building on this foundation, EXL recently unveiled a portfolio of agentic AI solutions, including enhancements to EXLerate.ai with autonomous agent-building capabilities, the launch of EXLdecision.ai for analytical model development, and domain-specific solutions such as EXL ClaimsAssist.ai for insurance claims management. These early pivots have helped us maintain strong growth.
Everyone is talking about GenAI. Where are you seeing real enterprise adoption?
The conversation has already moved from generative AI to agentic AI, and increasingly toward autonomous agents. Earlier, many clients were focused on proof-of-concept projects, but that phase is now largely behind us. Since the second half of 2025 and into early 2026, organisations have been accelerating toward full-scale enterprise adoption. As this shift takes hold, the focus has naturally moved from experimentation to outcomes, with return on investment becoming a central priority. Clients now expect implementation partners to deliver measurable business outcomes.
This requires two elements. First, the partner must deeply understand the client’s business domain, data environment, and operational workflows. Second, organisations need to be prepared for change management AI cannot simply be inserted into existing processes. Processes must be redesigned alongside AI deployment. When organisations embrace this transformation, AI can deliver strong ROI and far greater enterprise impact.
How is EXL integrating AI into its core offerings?
We approach AI adoption in three stages. First, we apply AI internally to enhance our own operations. Second, we embed AI within the operations we manage for clients. Third, we develop proprietary AI solutions that clients can deploy independently.
This is where EXLerate.ai plays a critical role. As the underlying technology framework, it enables enterprise-grade AI deployment by bringing together infrastructure layers, model orchestration, pre-built agents, and governance mechanisms.
It supports a multi-cloud environment across Azure, AWS, and Google Cloud, while remaining model-agnostic to optimise for cost, latency, and accuracy. With capabilities such as retrieval-augmented generation and prompt engineering, EXLerate.ai ensures that AI solutions are both scalable and grounded in client-specific data. Over the past year, this has been further strengthened through patents that advance our proprietary AI capabilities. For instance, we have got ten new patents in 2025.
What share of EXL’s revenue is AI-driven today?
In Q4 2025, our data, analytics, and AI business accounted for around 57% of total revenue. For the full year, it was approximately 55%, growing at roughly 18% annually. Our total addressable market has expanded significantly because we now participate not only in operations and analytics but also in data management and technology transformation. AI is unique in that business transformation and technology transformation must now happen together.
AI will continue to reshape roles, while also creating new opportunities across the enterprise. As organisations scale AI adoption, the focus is shifting toward building roles that combine domain expertise with data and AI capabilities. This is leading to the emergence of hybrid talent, professionals who can bridge business context with technological understanding, enabling organisations to drive more meaningful and sustained impact from AI.
Competition in the AI era will come from multiple directions, startups, traditional IT services firms, hyperscalers, AI companies like OpenAI and Anthropic, and consulting firms. As AI democratises intelligence, the real differentiator will be the ability to deliver business outcomes quickly. Organisations that combine operations expertise with analytics and AI capabilities are particularly well positioned. AI will gradually reduce simpler, repetitive tasks while creating new roles in prompt engineering, data annotation, AI governance, and responsible AI, giving rise to hybrid careers such as business technologists who understand both domains.
How is EXL reskilling its employees to work alongside AI systems?
We focus on three pathways. The first is enabling employees to work with AI, combining human ingenuity with AI capabilities. The second is enabling employees to work on AI, building tools and solutions. The third is enabling AI itself, which involves structuring and annotating data so that AI models can use it effectively. We offer structured training programmes and self-paced learning, and employees have access to sandbox environments where they can experiment with AI tools and build practical solutions.