How Telecom and Mining Companies Make AI Work in Production

The article opens with data on stalled enterprise AI (MIT: 95% of enterprise GenAI pilots show no P&L impact; Gartner: 60% of AI projects will be abandoned through 2026 over weak data foundations), arguing the root cause is data quality, not the models. TPG Telecom built a community-driven governance model that grew from 50 to over 300 members (150 active), assigning stewards to critical datasets and building a certified semantic layer via Alation's Semantic Model Mastering, shifting the internal question from "where did you get that data" to "are you using the certified source?" OceanaGold separately documented roughly 2,000 business metrics with rigorous definitions and ownership, which one leader credited when "the first time we threw Claude at it, first try, the discipline paid off completely." Both companies ran a "two-speed" model, delivering current AI use cases while building governance foundations in parallel.

Alation is a data-catalog and data-intelligence platform that tells a pharma team what data actually exists, who owns it, and whether it is trustworthy enough to support an AI use case.

This sits in the Design & Stress-Test stage of the AI use-case lifecycle: Named examples from the tools pharma teams use to scope, prototype and benchmark an AI use case before committing to build it — data-catalog, sandbox and evaluation platforms.