At the ET Edge CIO&Leader Annual Conference, Rucha Nanavati argued that AI transformation moves only as fast as the data foundation beneath it, and laid out what trustworthy enterprise data actually requires.
Rucha Nanavati, Chief – Farm Advanced Technologies at Mahindra & Mahindra, opened by connecting her session to the company’s RISE philosophy, before making her central argument: “speed is earned at the foundation.” The pace of any organisation’s AI transformation, she said, is set by the strength of its data and the effectiveness of its governance — not by which models it adopts.
Beyond general-purpose intelligence
Nanavati pushed back on the idea that generic, off-the-shelf AI delivers real enterprise value. Everyone, she noted, is offering some version of GPT; what enterprises actually need are highly contextual, domain-specific use cases — she pointed to Bloomberg GPT as an example of the kind of focused application that genuinely drives ROI, rather than broad, general-purpose intelligence.
The four pillars of great data
She then laid out what she called the non-negotiable qualities of enterprise data: it must be connected, integrated, accurate, comprehensive, and governed. “Data is the new oil,” she said, but only if it’s built right. Nanavati was direct about the cost of getting this wrong, noting that poor data quality carries a real, measurable price that percentage-based metrics often understate.
Comprehensiveness over volume
A key theme of her talk was resisting the temptation to equate more data with better understanding. Acquiring more customers, she argued, doesn’t automatically mean knowing them better. In what she called the era of hyper-personalisation, she warned enterprises: “Don’t fall for a single sliver of data or you’ll lose the understanding.” Depth and context, she said, matter more than sheer volume — comprehensiveness depends on the context an organisation is trying to build, not the size of its dataset.
Trust, lineage and privacy as strategic imperatives
Nanavati closed on a note of accountability. When an AI solution produces a wrong answer, she said, organisations need to be able to trace it; understanding data lineage and privacy is no longer a technical afterthought but central to trust. Her question to the room was blunt: “Do you trust your data?” For Nanavati, that trust — built on quality, traceability, and governance — is what will ultimately separate enterprises that scale AI successfully from those that don’t.