NTT DATA has internally defined the usage and deployment of agentic AI for two different purposes — to build custom agentic AI solutions for enterprises and to focus on developing robust AI-native businesses.
The company has established AIVista, a Silicon Valley-based wholly owned subsidiary of NTT DATA, to accelerate the launch and scaling of AI-native businesses. It plans to generate $2 billion in revenue from AI-native businesses by 2027.
Responding to a question on the company’s definition of agentic AI, Rajeev Singh, executive managing director, applications and business process services at NTT DATA, said: “NTT will build agentic AI, which are custom AI factories, for enterprises to help customers solve precise problems. Second, we are focusing on building robust AI-native businesses that will be infused into sectors like insurance, logistics, banking and contact centres, among others.”
AIVista, led by Bratin Saha as CEO, was formed on December 1, 2025.
Singh said the adoption level of AI in the manufacturing sector is still at an early stage in India. “The adoption level of AI and agentic AI in the manufacturing sector is pretty low and is at an early stage compared to North America. In the American region, we see companies creating digital twins and other technologies in manufacturing. I would be surprised if this adoption does not happen in India in the next one-year period.”
He also said that with the rapid evolution of AI solutions, the shelf life of technologies has reduced over time, and companies will have to continuously introduce new products to stay relevant.
“The issue of technical debt is increasingly being discussed, and with the swiftness with which AI is permeating organisations, it has become a challenge. If companies want to switch to an AI-driven world, it will not work with legacy technical debt. The way we see AI unfolding at the current stage, there is no other option but to adopt modern technologies to remain competitive in the new world. Companies will need to be agile when it comes to AI adoption,” he added.
Technical debt is a software development concept describing the long-term hidden costs of prioritising quick, suboptimal or temporary solutions over cleaner, more sustainable and well-designed code.
As the Business Process Outsourcing (BPO) industry faces potential job losses due to the rapid adoption of AI and agentic AI, Singh offered a different perspective, saying it will automate many sectors that were not previously part of the outsourcing ecosystem.
“Yes, there will be job losses in the BPO sector, but this also gives us an opportunity to move deeper into various segments that were not earlier part of BPO but have suddenly become part of the ecosystem as these models help automate processes. We have signed a deal with one company on AI-native architecture for BPO, wherein we will automate the business starting from customer experience to order entry to supply chain management. We have not done a business like this before,” he said.