The Revenue Integrity Gap: The Operating Model Comes First - Why AI Needs a Substrate Before It C...

In this episode, David sits down with Nicole Miller from Nordis Technologies (www.nordistechnologies.com)-a product-minded operational leader who spent eight years rebuilding the technology substrate of a 30-year-old company before ever pointing AI at it, to unpack what disciplined AI transformation actually looks like inside a legacy enterprise. Moving beyond the pressure to "just put AI on it," the conversation explores why intentional data architecture — not model selection — is the real prerequisite for AI ROI. Key topics covered: *The substrate-first approach:* While most companies rush to deploy AI on top of fragmented systems, Nicole spent years simplifying architecture, rewriting the intent of the technology stack, and building a patented platform before AI was ever in the picture. Her logic: you can't point AI at data that doesn't tell a story. AI-enabled vs. AI-centric vs. AI-native: A framework for understanding where your organization actually sits. Nicole argues there's no getting to AI-native in a thirty-year-old company without first being AI-enabled — and most leaders skip that step entirely. Technical debt as a strategic decision: Rather than chasing legacy debt indefinitely, Nicole made the hard call to draw a line — telling her technical team "we're not looking back" — and centered the organization on a new foundation built around data that tells a story (Espresso Hub). Data lakes as landfills: Without intentionality, data warehouses become dumping grounds of duplicative, incongruent information. Nicole's cautionary tale of a CTO who built a data warehouse under pressure — only to have it dismantled — illustrates what happens when governance and intent are absent. The falsehood of AI: AI alone won't solve anything. Without documented processes, defined systems of truth, and clear data definitions, AI just costs money and frustration. The real discipline is patience — understanding the business and the right data before letting AI work for you. Measuring ROI in people power: Nicole translated AI impact into FTE terms the leadership team could understand — "when AI is working for us, we don't increase the employee population by one, but we feel the lift of twenty." Concrete examples include coding productivity (every two engineers now contribute like three and a half) and customer service response times cut from 80 minutes to 10 via their AI agent, Ask Coral. The multi-year journey: A phased operational strategy — One Team One Dream → Ready Set Grow → Be the Spark → Aligned Execution → Unleashed Potential → New Frontiers → Intelligence Into Action — showing how bringing 180 people along takes longer but ensures longevity. Bringing the whole team: The hardest part wasn't the technology — it was educating the entire leadership team on AI. Nicole stresses that you can go anywhere fast by yourself, but taking an organization with you requires patience, inclusion, and a shared vision. The one decision: Define your outcome. Without a clearly defined outcome, the investment becomes impossible to measure and the measurement against it becomes impossible to track. The episode closes with Nicole's core message: AI on its own is not enough. You have to think about AI for your business — where you are in the story, then think about where AI fits. Discipline, patience, and intentionality are what separate organizations that build lasting AI capability from those that end up in the pilot graveyard.