In an interview with TimesTech, Sanjay Rohatgi, Senior Vice President and GM – APJ, Automation Anywhere, discusses how agentic AI is transforming enterprise automation from task-based execution to intelligent workflow orchestration. He shares insights on AI-driven business operations, the evolution of workforce roles, India’s growing leadership in automation adoption, enterprise-grade AI platforms, and the importance of embedding security and governance into scalable AI ecosystems.
Read the full interview here:
TimesTech: As enterprises move from experimenting with AI to deploying it at scale, how do you see agentic AI transforming day-to-day business operations across industries in the APJ region?
Sanjay: Agentic AI is fundamentally shifting how enterprises automate, moving from task-level execution to intelligent, end-to-end workflow orchestration. This is what we call Agentic Process Automation (APA). Driven by our Process Reasoning Engine (PRE), a model-agnostic platform securely orchestrates complex operations while delivering a significant improvement in workflow efficiency.
Across APJ, adoption maps to market maturity: Singapore is highly advanced, Australia and India are moving fast, and Japan is catching up promptly.
We are seeing strongest momentum in three verticals: BFSI, manufacturing, and public sector/utilities. In banking, for instance, agentic automation is transforming decision-heavy workflows like KYC, Anti-Money Laundering (AML), and loan processing.
TimesTech: There is growing discussion around AI augmenting human capabilities rather than replacing jobs. How do you see enterprise roles evolving as AI agents take over repetitive and process-heavy tasks?
Sanjay: AI has massive potential to augment enterprise roles. It can empower employees to focus on high-value problem-solving. Today, workflows are designed with a hybrid approach: traditional Robotic Process Automation (RPA) handles the deterministic, rule-based tasks; agentic AI manages the probabilistic tasks where contextual reasoning is needed; and humans stay in the loop for the critical decision-making. While AI can dramatically accelerate analysis and reduce errors, human oversight ensures trusted guardrails, compliance, and governance remain completely intact.
TimesTech: India is rapidly emerging as a key market for AI-led automation adoption. What factors are driving this momentum, and how are Indian enterprises approaching agentic AI differently from global markets?
Sanjay: Enterprises everywhere are working with similar AI models and asking the same fundamental question: what tangible business value can AI deliver? They are focused on identical metrics: process efficiency, accelerated time-to-market, and strict, outcome-based pricing return on investment (ROI).
Where India stands out is its massive concentration of conglomerates, BFSI, and Global Capability Centres (GCCs). While the government’s India AI Mission provides an excellent structural push, India’s private sector is aggressively leading the charge in deploying scaled, outcome-driven agentic automation faster than global peers.
TimesTech: Traditional automation often operated in silos. How is the shift toward orchestrated, end-to-end workflows changing enterprise expectations from automation platforms today?
Sanjay: Enterprise expectations have shifted from isolated automation projects to cross-functional workflow orchestration. Siloed agents within a single application cannot solve enterprise realities. The entirety of a process comprising multiple tasks needs to be orchestrated. Enterprise focus is also on platform scalability, interoperability, and measurable business outcomes.
This shift is fundamentally changing how enterprises evaluate automation and AI investments.
A horizontal orchestration platform becomes important in such a scenario. Such a platform paves way for the tying together of disparate systems into a single unified execution layer, delivering scalability and governance instead of fragmented tech silos.
TimesTech: With several players entering the AI automation space, what differentiates a truly enterprise-grade agentic process automation platform in terms of scalability, integration, and workflow intelligence?
Sanjay: A truly enterprise-grade platform is defined by how securely and effectively it orchestrates agents at scale.
The barriers to building a basic AI agent are incredibly low, which is driving a dangerous trend toward ‘agent sprawl,’ meaning, hundreds of disconnected, ungoverned agents creating chaos across an organization. This makes the build-versus-buy debate critical.
A platform-led approach replaces fragmentation with centralized governance, security, and auditability. What differentiates an enterprise-grade APA platform is its ability to solve the most complex, cross-enterprise processes using agentic process automation. This must be done in a manner that is uniquely secure, simple, scalable, sustainable, and smart.
TimesTech: As AI adoption accelerates, AI-driven cyber threats are also becoming more sophisticated. How do you see enterprises balancing innovation with security, and what role can AI itself play in strengthening cybersecurity resilience?
Sanjay: Cybersecurity is a foundational pillar of agentic AI for the enterprise, especially as threat actors begin leveraging AI to scale their own exploits.
To protect the expanded enterprise attack surface, guardrails cannot be an afterthought, rather embedded across all three layers of the AI stack: the compute/GPU infrastructure, the LLM models, and the applied AI workflow layer.
By embedding automated security, auditability, and role-based access directly into the workflow architecture, enterprises can innovate without sacrificing compliance.
