Making AI Pay Off at Enterprise Scale: Insights from MachineCon NY

For all the talk of AI revolutionizing business, a statistic remains stubbornly telling: only about 1 in 10 AI proof-of-concept projects actually scales to enterprise value. The promise is there: operational efficiency and predictive insights, but most companies remain stuck in experimentation. So what separates fleeting experiments from true transformation?

The gap, as it turns out, is organizational, cultural, and strategic. The task confronting many leadership teams today is whether they can make it work for their company at scale, and whether the investment will justify itself.

In a recent panel at MachineCon New York, moderated by Asif Ghatala, Vice President and Business Unit Head at Tiger Analytics, this challenge was unpacked by senior leaders with firsthand experience navigating that journey. The speakers included Arys Nogueron, Campaign Director of Data Analytics, Reporting, and Governance at NYU Langone Health; Manav Misra, Chief Data and Analytics Officer at Regions Bank; and Rama Donepudi, SVP and Global CIO at Mead Johnson Nutrition. Together, they represented a cross-section of industries: finance, healthcare, and consumer nutrition, but voiced common themes about how to unlock AI’s true ROI.

From Concept to Confidence

Getting past the POC stage takes organizational trust, speed, and a willingness to fail productively. Several noted that early AI efforts were fast and scrappy, with short feedback loops, numerous failures, and small wins that helped build credibility. In regulated industries like healthcare and financial services, even small experiments often require diplomatic stakeholder engagement and stringent risk assessments.