Redefining Work: The Rise of Agentic AI in the Workplace

.Companies across sectors are beginning to deploy agentic AI systems – software agents capable of making decisions, executing tasks, and collaborating with humans at work. And as this shift gathers pace, business leaders are rethinking not just technology strategy, but also how jobs are designed, skills are built, and teams are managed.Last week, we had an insightful discussion on this.From tools to co-workersSridhar Hariharasubramanian, senior director at Salesforce India, said the rise of agentic AI represents a structural shift. “What we are seeing is a transition from AI as a passive tool to AI as an active collaborator,” he said. “That fundamentally changes how work gets done.” However, Sridhar cautioned that this transition is not just a technology upgrade – it is a people transformation.IPL Auction 2026IPL Auction 2026: Full list of sold and unsold players for all teamsIPL 2026 team and squad List: Updated players for all 10 Teams; who got whomAs AI agents take on routine and even complex tasks, human roles will increasingly centre on creative problemsolving, strategic thinking, and collaboration.A surprisingly critical technical skill, he added, is what many now call prompt engineering – or, more accurately, context engineering. “The quality of outcomes you get from AI is directly proportional to the quality of the questions you ask,” he said. In an agentic enterprise, knowing how to guide, frame, and supervise AI becomes as important as knowing how to code.Sridhar emphasised the importance of continuously learning, unlearning, and relearning, noting that this will define career resilience in the AI age.At Salesforce, he said, employees are actively encouraged to experiment with AI tools so they are comfortable working alongside digital agents as co-workers rather than viewing them as threats.Building adaptability at scaleThat emphasis on learning was also emphasised by Sumati Mohan, director at Cognizant India.At Cognizant, Sumati said, the response to AI has been to invest deeply in structured upskilling and cross-skilling programmes that span not just engineers, but also consultants, sales teams, and operational staff. These efforts are tracked through internal academies to ensure that learning translates into deployable capability.The company has also set up dedicated Agentforce labs and cohort-based learning initiatives to accelerate hands-on exposure.Gen Z employees, Sumati said, have emerged as some of the fastest adopters of agentic AI, quickly experimenting with new ways of working. Mid-level managers, often seen as the hardest group to shift, are also adapting faster than expected. “The demand from clients is very clear,” she said. “They want responsible, outcome-driven AI. And that means our people need to be ready – not just technically, but ethically and operationally.”.Beyond the digital native firmsWhile much of the early AI narrative has focused on digital-native companies, Sanjeev Malhotra, CEO of Nasscom Centre of Excellence for IoT and AI, offered a broader view from his work across India’s innovation ecosystem. “Adoption is happening everywhere,” he said. “Yes, digital natives are ahead, but even companies that were never digital are now embracing agentic AI.” The Nasscom centre works closely with startups, enterprises, govt bodies, and SMEs.What Sanjeev is seeing, particularly among small and medium enterprises, is a sharp focus on returns on investment. “SMEs don’t experiment for the sake of experimentation,” he said. “They adopt when they see value – and now they are seeing it.” From automating quotations and procurement processes to handling customer queries and predictive maintenance, agentic AI is delivering tangible productivity gains. He added that collaboration between enterprises and startups is accelerating these innovations.From pilots to productionIf agentic AI is moving beyond experimentation, telecom companies are a prime example – and Shreeya Rashinkar, head of global solutions and presales at Tech Mahindra, has a front-row seat to that transition. Over the past year, Shreeya noted, there has been a marked shift from pilots to production deployments. “Pilots are easy,” she said. “Scaling is hard.” Moving into production requires rigorous testing, strong data foundations, governance frameworks, and regional compliance – particularly for global telcos operating across jurisdictions.New testing and governance capabilities within platforms like Salesforce’s Agentforce are helping accelerate this transition, making large-scale deployment more feasible. But Shreeya noted that every process does not need AI, and not every AI use case delivers RoI. The real work, she said, lies in distinguishing between simple automation, AI-driven intelligence, and tasks that should remain human-led. Tech Mahindra helps clients make those distinctions.Redefining work, not replacing itAcross the discussion, a shared theme emerged: agentic AI is not about replacing humans, but about redefining work. Roles will change. Some tasks will disappear, others will emerge, and many will be reshaped.