The Hard Truth About Getting Your Product Team Ready for AI

You’ve sent your product managers to AI certification courses. You’ve subscribed to ChatGPT Enterprise. You’ve hired a few data scientists. So why are your AI product initiatives still stalling after the first sprint?

Because getting your product team ready for AI isn’t a training problem—it’s a fundamental rethinking of how product organizations work. Models drift. Data changes constantly. What counts as a “bug” becomes ambiguous. Your traditional PM playbook doesn’t apply, and most companies are discovering this the hard way, months into failed AI initiatives.

In this episode of The Accelerate Podcast, host Becky Flint (CEO & Founder, Dragonboat) sits down with Bhaskar Deka, GVP of Product Management – AI & Learner Experience at Skillsoft and author of “Infused: Building AI-Infused Products & Services.” With experience leading product organizations through three startups to successful exits and working with enterprises like Informatica, and Cornerstone, Bhaskar gets real about what AI readiness actually requires and why most companies are approaching it wrong.

What you’ll hear:

• The “theoretical minimum” for AI PMs: The DEMO framework (Data, Explainability, Ops, Models) and why traditional product craft skills aren’t enough for AI products.
• Three types of AI products: AI-enhanced vs. AI-core vs. AI-infused: where your product actually fits, what that means for architecture, and how it changes your staffing and portfolio decisions.
• Why deterministic thinking fails: Moving from bugs to model drift, from static behavior to probabilistic outcomes, and the new metrics you actually need.
• The visibility crisis: How hidden dependencies and backend constraints block AI progress, and why one company’s “UI problem” was really a multi-year architecture issue nobody could see.
• Build, buy, or partner for AI: Making smart portfolio decisions when competitors are acquiring AI-native startups, and you’re resource-constrained.
• Low-hanging fruit that actually works: Where to start AI infusion: analytics, forecasting, workflow coaching, and why studying your existing workflows matters more than rushing to production.
• Signal vs. noise in evolving models: How great product leaders separate data that should change strategy from data they should ignore as AI products continuously learn.

This isn’t about AI hype or transformation success stories—it’s about the operational reality of preparing product teams for a fundamentally different type of product, the portfolio complexity it creates, and why spreadsheets and good intentions aren’t enough.

Reference

• Bhaskar Deka on LinkedIn: https://www.linkedin.com/in/bhaskar-deka-62a4171/
• Becky Flint on LinkedIn: https://www.linkedin.com/in/beckyflint/