Ep. 97 - Russell Long: Stop Blaming AI. Your Data is the Problem

Every CIO I talk to wants to know how to get AI ready. But according to Russell Long, co-founder and CEO of Authentica Solutions, most institutions are asking the wrong question. The real bottleneck isn't choosing the right AI model—it's unifying the data that feeds it.

In this episode, host Jeff Dillon sits down with Russell, who has spent two decades building the data infrastructure that sits quietly underneath education technology. He's taken a company through acquisition by BrightBytes (later acquired by Google), seen his DataSense platform find a second home inside Microsoft, and rebuilt Authentica around a modern education intelligence platform backed by a $6.2 million Series C round.

Russell shares hard-won lessons from the classroom, the enterprise, and global scale. He explains why "technical turn-on" doesn't equal adoption, why universities need to re-evaluate who they actually serve, and how a state-level nursing shortage was really a data problem in disguise. He also introduces the Adoption Improvement Method (AIM)—a framework for ensuring new technology actually gets used, not just purchased.

For anyone leading digital strategy, data governance, or AI adoption in higher ed, this episode is a grounded, practical look at what it actually takes to build the foundation for intelligent systems.

Key Takeaways

AI Readiness Starts with Data Unification, Not the Model: You can't trust AI answers if you don't trust the data behind them. Institutions need to unify their data into a "gold layer" of trusted information before they can confidently explore AI. Garbage in, garbage out—at scale.

Technical Turn-On Is Not Adoption: Just because a system is available doesn't mean it's being used. Many institutions confuse "turned on" with "adopted." Real adoption requires consistency, training, and ongoing engagement—not just a contract signature.

The Adoption Improvement Method (AIM) Is a Framework for Success: Authentica's AIM methodology includes 11 components for implementing enterprise technology, measuring impact, and revisiting change management. If adoption is low, it may be the wrong tech, the wrong onboarding, or just a bad fit.

Bad Data Leads to Bad Policy: Russell shares the story of a state nursing program that thought it was nearly full—until unified data revealed it was short by 2,000 spots annually and that 72% of nursing students were dropping out by sophomore year, not the assumed 36%. Acting on assumed data leads to bad decisions.

Data B