What does it actually take to move from AI adoption to AI transformation? It starts with a structural decision most organizations haven't made: giving someone AI transformation as their only job.

Alex Pugh, director of finance AI operations at Insight, sits inside exactly that model — a dedicated AI ops team embedded within a single business function.

This conversation gets into the accountability structures, the hard-won use case lessons, and the one metric that tells you transformation has worked (it's not what you think).

Request a meeting with an Insight AI specialist to get the right framework for accountability structures that turn AI investment into measurable business outcomes: https://www.insight.com/en_US/what-we-do/expertise/data-and-ai.html

Also worth your time: How To Shift From Reactive to Proactive IT Operations: https://www.insight.com/en_US/content-and-resources/blog/how-to-shift-from-reactive-to-proactive-it-operations.html

And, why prompting isn't the path to AI ROI: https://www.insight.com/en_US/content-and-resources/insight-on/the-product-manager-mindset-why-prompting-isnt-the-path-to-ai-roi.html

If you're still in adoption mode and feeling stuck, here's how BlackLine successfully hit 99% Gemini adoption: https://www.insight.com/en_US/content-and-resources/case-studies/blackline-hits-99-adoption-of-gemini-enterprise-with-insight.html

Subscribe for more conversations on AI strategy, IT modernization, and the decisions that separate leaders from the rest.

#AItransformation #EnterpriseAI #InsightOn Chapters

00:00 — Welcome and introduction

05:36 — Why domain expertise beats pure AI expertise

08:02 — Dedicated AI team vs. distributed champion model

12:56 — Meet the Thundercats — Insight's dedicated finance AI ops team

16:07 — AI adoption vs. AI transformation — there's a difference

22:17 — The CFO's mission and the case for going all-in

25:20 — The bitter lesson and why waiting on AI is not a strategy

28:39 — How to find real AI use cases inside your organization

34:03 — Nine out of 10 use cases are not AI opportunities

46:26 — Rapid fire — worst advice, best metrics, and how you know it's working