‘The Internet of Agents is the next shift’: Wipro tells us why the next wave of AI agents could boost your business more than ever
Share this article
Join the conversation
As AI and agents become a common sight in businesses of all sizes, the impetus now is to try and harness the benefits this technology can bring.
Networks of autonomous AI agents are now able to discover information, collaborate, negotiate and execute tasks on behalf of people and organizations, representing a shift from an internet built around websites and applications to one powered by an intelligent digital ecosystem.
We spoke to Hari Shetty, Chief Strategist and Technology Officer at Wipro, to find out more on this “Internet of Agents”.
Latest Videos FromTechRadar
- How is AI adoption moving from isolated pilots to enterprise-scale ecosystems?
For a long time, enterprise AI adoption looked like a series of disconnected experiments: a chatbot here, a copilot there, a proof of concept that never quite made it into production. According to industry reports, that phase is ending, with 40% of large organisations now actively scaling agentic AI. What is emerging instead is an ecosystem of deliberately connected platforms, people, and processes.
Three distinct kinds of change are happening at once. There is cost transformation, which is really about efficiency and productivity. There is technology transformation, which is the harder engineering work underneath, building the data models and governance that everything else depends on. And there is business transformation, which is the most ambitious of the three: taking a process apart and rebuilding it around a business outcome the organisation is willing to commit to.
Most industry conversations still centre on cost transformation, because it is the easiest to measure. But that is the trap: layering agents onto existing processes for marginal gains, while missing the bigger opportunity beyond it. Business transformation is harder to demonstrate and sell, but it is where the real value lives. This is the shift we are pursuing through Wipro Intelligence: connecting data, processes, platforms and agents to move clients beyond productivity improvements and one-off experiments towards measurable business outcomes.
- How will AI agents collaborate across organisational, geographic and technological boundaries?
AI agents will collaborate through orchestration, with specialised agents working together across models, data sources and enterprise systems to complete complex workflows. Understanding the evolving client requirements, we recently introduced a dedicated AI-Native Business and Platforms Unit. The unit focuses on building these agents and the orchestration layer that enables each agent to identify the right capability, access the right information and coordinate its actions across the enterprise.
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
Underneath this will sit a distributed architecture combining frontier models with smaller, enterprise-specific models built around an organisation’s own data and use cases. Most enterprise tasks do not need a frontier model; a smaller model grounded in the right data is often more efficient, easier to govern and better suited to the domain. The orchestration layer enables agents to work across this mix, selecting the model, data and tools best suited to each step in the workflow.
Collaboration across organisations, geographies and technology platforms will require more than connectivity. Agents will need verifiable identities, clearly defined permissions and reliable ways to discover and trust one another. Equally important is a control plane that allows people to set the level of autonomy, intervene when necessary and review every action taken by an agent. This gives enterprises the transparency and auditability to scale agentic workflows without weakening control over data, access or accountability.
- What does the Internet of Agents mean in practice, and how does it change enterprise computing?
Historically, the internet connected people to information. The Internet of Agents is the next shift: connecting AI-driven intelligence to enterprise systems, workflows and other agents, so software can act on information rather than simply present it. That changes the purpose of enterprise computing, moving it from reactive towards autonomous.
In practice, there are two broad classes of agent. Assistant agents work alongside individuals, helping them summarise emails, analyse information or develop and test software. More significant are agents embedded directly into business processes. These often operate invisibly, moving across systems to resolve an issue, complete a transaction or orchestrate a workflow, escalating only the decisions that need human judgement.
Given the level of hype surrounding agentic AI, proof over promise matters more than ever. The starting point should be the industry’s core processes and the underlying KPIs that directly affect business performance. Those processes can then be revisited with AI-first thinking to determine how they should be reimagined and which outcomes the enterprise is prepared to commit to. This outcome-led approach also helps protect clients from failed investments and underperformance, because value is measured against the business result rather than the number of agents deployed.
- What use cases are emerging from this shift towards agentic ecosystems, and how can IT services companies catalyse adoption of the Internet of Agents?
Agentic ecosystems are already reshaping work across the enterprise, from customer onboarding and supply chain investigations to business reporting, financial planning, marketing, employee services and enterprise risk management. This shows up first in processes spanning multiple systems and decisions: agents supporting financial reconciliation, fraud analysis and invoice workflows, with applications in logistics and industry ranging from fleet coordination to plant maintenance and drilling efficiency.
These use cases go beyond accelerating isolated tasks. They involve redesigning end-to-end processes around agents that use enterprise data, apply business rules and coordinate action across systems, requiring a stronger operational foundation than individual AI deployments.
Enterprises need a layer that connects agents to the right models and trusted data, integrates them with core systems, and provides clear controls over how they operate. We address this through a portfolio of AI-first solutions spanning enterprise functions and industry value chains. Horizontal solutions combine human-agent capabilities across HR, finance and supply chain, while industry-specific solutions target mission-critical processes such as KYC in banking, asset management in manufacturing, energy and utilities, and autonomous networks in telecommunications. By bringing together domain expertise, AI and automation, these solutions are designed to deliver measurable business outcomes rather than add another layer of technology.
- What operational challenges must be solved before agentic AI can scale?
Agentic AI will only scale when enterprises can trust agents under real operating conditions. Moving beyond controlled pilots means handling business volumes and unexpected situations without compromising accuracy or consistency.
Trust also depends on traceability. Enterprises need to be able to audit how decisions were reached, understand the work performed and recover when processes fail. Resilience and appropriate human oversight must therefore be built into agents from the outset, particularly where business or regulatory risks are higher.
Ownership matters too. To scale agentic AI, enterprises need operating models that give agents a defined role in live workflows while keeping people accountable for the outcomes. Human owners must be able to set autonomy levels, steer behaviour, review performance and intervene when needed, supported by clear audit trails and business-linked measures.
- Why will future enterprise success depend on trusted operating models between people and agents, rather than agent counts?
Agent counts reveal little about enterprise progress, because there is no consistent way to define them. One organisation might describe ten functions as ten separate agents; another might bundle the same capability into one. A better measure is the work people and agents accomplish together, judged through business outcomes: more effective fraud detection, faster processing, or more responsive operations.
This requires AI fluency spread across the organisation, not concentrated in specialist teams. People at every level need to understand how to work with agents, when to rely on them, and where human judgement remains essential. Leaders play a significant role in creating psychologically safe environments where teams can experiment with technology.
According to a recent industry report, only 13% of employees feel rewarded or supported for reinventing their work with AI when the results aren’t immediate. That challenge reflects the J-curve of AI value: investment comes first, followed by a period in which benefits can appear limited or difficult to measure. Leaders need to sustain confidence through this phase, because the early gains are incremental and the larger return emerges only as more of the value stream moves to the new operating model. This makes agentic AI as much an organisational and cultural change as a technical one.
The enterprises that succeed will be those that build enough trust for people and agents to work together confidently, keeping accountability clear even as more work becomes autonomous.

Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
Source
Share this article Join the conversation Follow us Add us as a preferred source on Google Newsletter Subscribe to our newsletter As AI and agents become a common sight in businesses of all sizes, the impetus now is to try and harness the benefits this technology can bring. Networks of…
Recent Posts
- How to watch Lauren Price vs Michaela Kotaskova: live stream boxing for FREE
- How to watch Lauren Price vs Michaela Kotaskova: live stream boxing for FREE
- ExpressVPN undercuts its own Black Friday deal with its low cost VPN plan now cheaper and better value than NordVPN, with secure AI, email aliasing, ad, tracker and malware blocking all for $2.79
- UN court cites military Facebook posts while finding Myanmar guilty of genocide
- Ray-Ban Meta (Gen 3) review: Minor upgrades that cost too much
Archives
- November 2026
- October 2026
- September 2026
- August 2026
- July 2026
- June 2026
- May 2026
- April 2026
- March 2026
- February 2026
- January 2026
- December 2025
- November 2025
- October 2025
- September 2025
- August 2025
- July 2025
- June 2025
- May 2025
- April 2025
- March 2025
- February 2025
- January 2025
- December 2024
- November 2024
- October 2024
- September 2024
- August 2024
- July 2024
- June 2024
- May 2024
- April 2024
- March 2024
- February 2024
- January 2024
- December 2023
- November 2023