Technology Leaders Link AI Pilots To P&L

Economic Times CIO reported that at the ETCIO Annual Conclave 2026 in Goa a panel titled "From Pilot to P&L: Why Some AI Bets Scale and Others Quietly Die" discussed enterprise AI adoption. The session was moderated by Sneha Jha, Editor, ETCIO, and included Rakesh Bhardwaj, Global CIO, Lupin; Poorav Sheth, CDIO, Hindalco; and Hitesh Shah, Vice President, Manufacturing Business, Kyndryl India. "Making a demo work can be done with a small set of data and a narrow scope, but enterprise reality is very different," Bhardwaj said, per Economic Times CIO. The outlet also reported Sheth saying, "AI value in manufacturing has to show up in business results such as reduced downtime, better safety, stronger resilience and improved margins."

Companies that move beyond narrow proofs-of-concept typically confront three technical barriers: integrating AI with core systems, operating on validated enterprise data environments, and meeting security and privacy controls. Industry-pattern observations: integration work often dominates engineering effort once a pilot leaves a sandbox, and explainability tooling becomes essential where business users must trust recommendations before operationalizing them. Observed patterns in comparable deployments: analytics platforms and explainability features are recurring enablers of adoption in regulated and operationally sensitive domains such as manufacturing.

Tying pilots to measurable P&L and operational KPIs reduces the ambiguity evaluators face when deciding whether to fund scale-up investments. Industry observers note that when pilots produce clear impacts on downtime, safety or margins, procurement and operations stakeholders are more likely to sponsor wider rollouts. For practitioners: framing project metrics as operational outcomes instead of narrow model metrics (for example, reduction in mean time to repair rather than model accuracy alone) changes which data sources, validation steps and governance controls are prioritized.

Indicators an observer should follow include:

• •whether pilot metrics are reported as business KPIs rather than model-only metrics
• •the presence of cross-functional ownership or a platform team responsible for integration and governance
• •investments in explainability and validation workflows that let business users audit recommendations. Industry context: conferences and practitioner forums will likely continue to surface case studies that highlight these practical integration and adoption challenges