New Delhi, Aug 12:The next challenge for artificial intelligence may not be smarter models, but whether enterprises are ready to deploy AI effectively at scale, according to Xebia.
While AI technology has advanced rapidly, enterprises are still working to move initiatives from experimentation and pilots into production. Challenges around data, legacy systems, governance, infrastructure and organisational readiness continue to affect adoption.
“Technology is ready for the next revolution of AI. The question is whether we as humans and organisations are ready for that,” said Mayank Verma, Global Head – Data and AI at Xebia. According to Verma, successful AI deployment requires a clearly defined outcome, ownership, production-grade infrastructure and governance, along with the willingness within organisations to change existing processes.
Data remains a key part of the challenge. While enterprises have traditionally relied on structured data, AI increasingly requires broader enterprise context, including documents, policies, meeting notes, audio, video and other information.
“AI requires not only traditional data but also enterprise context,” Verma said. The issue becomes more important as enterprises move from copilots towards autonomous AI agents. Businesses will need to determine where AI can operate independently, where human oversight is required and what information agents need to make effective decisions.
Xebia is addressing this through Xebia Axis, its agentic data foundation, which is designed to help enterprises unify fragmented data, modernise legacy systems and prepare enterprise data for AI agents.
The platform can also monitor data pipelines and help identify potential causes of failures, allowing human teams to focus on corrective action.
As AI adoption expands, enterprises will also need to look beyond token consumption and consider the overall economics and business value of AI workloads. Different models can be used depending on the complexity and requirements of individual use cases.
The next phase of enterprise AI, therefore, is likely to depend as much on enterprise readiness particularly data, infrastructure and governance as it does on advances in AI models.