Artificial Intelligence has transitioned from a stage of experimentation to an essential requirement for businesses in just four years. Following the generative AI surge of 2022, companies have invested billions in models, infrastructure, and AI-driven strategies. However, a challenging reality remains: most organizations still struggle to achieve measurable returns on their investments.
On a recent episode of the Analytics Insight Podcast, host Priya Dialani spoke with Abhijeet Vaidya, Head of Data and AI Practice at Emids, about what he calls the growing ‘Gen-AI divide.’ While adoption metrics soar, studies suggest that the vast majority of enterprise AI projects remain stuck in the pilot stage or fail to deliver clear ROI.
“AI doesn’t fail because models aren’t powerful,” the Head of Data and AI Practice at Emids comments. “It fails when context is missing.”
Healthcare illustrates the challenge sharply. The sector relies on vast volumes of unstructured data, including clinical notes, lab reports, and imaging results, all layered within strict regulatory systems. Implementing AI requires technical skills and the ability to understand specific work and to establish complete operational systems.
Abhijeet provided a demonstration of a healthcare network that used AI technology to categorize patient complaints. The system reduced processing time from weeks to minutes. Yet the downstream legacy workflow could not integrate AI outputs. The technology worked. The return did not materialize.
“You cannot deploy AI in isolation,” he noted. “Enterprises operate within ecosystems built over decades.”
At the same time, India is emerging as a compelling testbed for healthcare AI. With its scale, digital public infrastructure, and expanding talent pool, AI-driven care planning and remote diagnostics are already shortening consultation times and improving access in rural areas.
Abhijeet needs to guide his organization through AI development by creating operational systems. The hype around technology will produce lasting results only when organizations combine their technological systems with their real-world operations, regulatory needs, and business performance targets.