The world is experimenting with a transition from generative AI to agentic AI—one that could fundamentally reshape artificial intelligence. This shift marks an evolution from reactive, “armchair” systems that primarily generate content in response to human prompts to proactive, autonomous “co-worker” agents capable of multi-step reasoning, tool integration, persistent memory, planning, and independent goal execution with minimal oversight.
But is India—and its existing AI infrastructure—ready for this transformation?
THE WEEK speaks to Siddharth Dhar, President & Global Head – Digital IT Operations & AI at Hexaware Technologies, which recently launched Agentverse, an enterprise AI agent platform featuring over 600 ready-to-deploy AI agents designed to help organisations operationalise agentic AI across business and technology functions. Excerpts:
Q/ How would you assess India’s current cloud infrastructure maturity in enabling widespread deployment of agentic AI systems?
A/ India has reached a point where cloud infrastructure is no longer the primary constraint. With strong hyperscaler presence and growing enterprise adoption, access is largely in place. The real question now is how effectively enterprises are using that infrastructure for AI workloads. We are moving from a phase of infrastructure readiness to execution readiness, where the focus is on integrating AI into business processes rather than just building capability.
Q/ What are the biggest challenges in India’s cloud readiness for scaling agentic AI?
A/ The challenges are less about availability and more about how enterprise environments are structured. Many organisations still operate with fragmented systems and limited standardisation. Fragmented legacy systems, poor data quality, and inadequate governance remain key gaps, and without these, scaling AI becomes difficult. Beyond enterprise readiness, infrastructure gaps persist, like unreliable power supply, lack of high-density AI-ready data centres, and limited fibre connectivity outside metros continue to constrain scale. Many AI startups also lack a sustainable compute strategy, and operational and legal control over data remains an unresolved challenge. So, the issue is not cloud access, but how well enterprises can operationalise AI on top of existing architectures.
Q/ How prepared is India’s cloud ecosystem to handle security and governance risks of agentic AI?
A/ The ecosystem is evolving, but agentic AI introduces a new level of complexity because systems are not just responding but taking actions. This increases the importance of identity, access control, and continuous monitoring. Enterprises need to move from static security models to more dynamic, real-time governance frameworks. This is where AI-led cybersecurity and identity management become critical, ensuring that autonomy does not come at the cost of control.
Q/ How do India’s digital public infrastructure developments support agentic AI innovation?
A/ India’s digital public infrastructure has created a strong foundation for scalable digital systems by enabling interoperability and large-scale adoption. Platforms like Aadhaar, UPI, and ONDC show how systems can work together at scale. For agentic AI, this kind of infrastructure is important because it supports integration across services.
On the development side, the IndiaAI Mission’s investment in a national compute facility, open dataset platforms like AIKosh, and indigenous foundation models such as Sarvam AI and BharatGen AI are directly enabling AI innovation. Real-world deployments like AI-powered crowd management at Mahakumbh 2025 and the Bhashini chatbot for multilingual assistance further demonstrate how DPI platforms are already supporting agentic applications at scale. That said, flexibility across cloud environments and reducing vendor lock-in will be important as enterprises look to build more independent and portable AI systems.
Q/ What key enablers are needed for India to lead in the agentic AI internet?
A/ The focus needs to be on execution. This includes strengthening cloud and data infrastructure, building clear governance frameworks, and accelerating enterprise adoption. At the same time, there needs to be a shift in how we think about talent — from general AI awareness to deeper engineering capability. If these elements come together, India is well positioned to play a leading role.
Q/ While India has the world’s largest pool of “AI-literate” developers, only a fraction are “AI engineers”. Why is this so?
A/ There is a clear difference between understanding AI tools and building AI systems. A recent report highlights this starkly. Nearly 90 per cent of Indian engineers feel AI-ready, but only less than 20 per cent actually are.
Many developers today are familiar with AI concepts and platforms, but fewer have experience in designing, deploying, and scaling these systems in real-world environments.
The difficulty in filling AI roles is most acute when it comes to practical, system-building skills. AI talent is also largely concentrated in metros, and women represent only 36 per cent of GenAI learners—gaps that point to the need for more inclusive, distributed skilling efforts.
Q/ Does India have enough specialised AI engineering talent, and what is needed to accelerate it?
A/ The talent exists and is growing but not yet at the scale required for complex, enterprise-grade AI systems. What’s needed is more focus on real-world problem solving, system design, and integration. Enterprises also have a role to play by creating environments where engineers can learn through live deployments rather than just training programmes.
At Hexaware, our approach has been to embed AI into every role rather than treat it as a specialist function. This has led to 90 per cent of our employees becoming AI-certified, not through mandates, but by building an ecosystem that makes learning continuous and contextual. This helps us accelerate the depth of AI capability at scale.
Q/ Tell us about your Agentverse. How does Hexaware’s Agentverse help enterprises move from AI pilots and experimentation to large-scale deployment?
A/ Most enterprises today are not struggling with AI capability but with scaling it beyond isolated pilots. The challenge is that these pilots often sit outside core systems and don’t translate into day-to-day operations. Agentverse addresses this by bringing orchestration, integration, and governance together on one platform. With ready-to-deploy agents that work within existing enterprise systems, it allows organisations to move from experimenting with AI to actually running workflows on it. This is very much in line with what we’re seeing across the industry—a clear shift from pilots to production.
Q/ Which business functions and operations is Agentverse designed to improve, and what governance frameworks or policy guardrails are in place?
A/ Agentverse is designed to improve both technology operations and business functions. On the IT side, it supports development, testing, cloud, and ongoing operations. On the business side, it helps streamline customer experience, finance, compliance, and industry-specific workflows like underwriting or claims.
From a governance standpoint, it is built with enterprise controls at the core. That includes role-based access, audit trails, policy enforcement, and observability. So, agents are not just executing tasks, they are doing so within clearly defined guardrails that align with enterprise security, compliance, and accountability requirements.