This highlight video is from the May 22, 2026 live discussion Design Charrette | AI In The Loop Coaching

"we've been in this flurry of data collection, to what end?" — Nicole DeFalco, Upsurge Advisors

The real work on the table: People leaders are spending heavily on learning but can't prove what it's moving. Anton Maletich (Alter Domus) brought a problem he's seen at three companies: the LMS captures activity but doesn't surface insight, managers can't see where their teams stand, and there's no centralized reporting across programs. It matters now because boards are already pricing L&D against AI-driven alternatives — and L&D without proof points is the easy cut.

Who was in the room: Anton Maletich (Alter Domus) as Internal Client, Kyle Swanson (Ziel) as AI Strategist, David Scherer (ELE) as Skeptic, Matt Sernaker (Lake Forest Center for Leadership) as Executive Producer — plus senior L&D, talent, and people leaders from Upsurge Advisors, UL Solutions, LAK Group, Positive Delta, Robertshaw, National Louis University, and others.

Clean the data before you point AI at it

"you don't have clean data. Everything you build on top of it is worthless." — Anton Maletich, Alter Domus

Duplicate profiles, blank fields, employees on leave — they quietly distort every completion number you report up. Anton's first AI use case wasn't a dashboard. It was an audit of the data itself.

Lead with the business question, not the LMS report

"what are the questions you're really asking of the data?" — Kyle Swanson, Ziel

If the business doesn't already measure the outcome, learning shouldn't invent a metric for it. The AI question comes third — after the business question and the data question.

Completion is not performance

"Does it matter if the training was completed, if the performance has moved." — Adrienne Guerrero, Positive Delta

Adrienne's group surfaced the sharper question: maybe the training didn't move the needle at all. Manager influence, peer coaching, and informal learning often do more. Measure the performance change first. Then back into what actually caused it.

The boardroom signal

Vilson Simon (LAK Group) shared a story that landed hard. A $200K L&D vendor was replaced by a $50K AI-powered platform — surfaced by two IT folks who saw the savings before L&D had built its case. "There's a lot of noise today in boardrooms," he said, "where we can cut costs across the organization if it's using AI." If L&D can't quantify its value, someone else will quantify the alternative.

That is the organizational intelligence shift: business data may show what changed, but people intelligence helps leaders understand what the work is becoming.

---

What to try next

1. Run an AI-driven data audit before any reporting build. Duplicates, blanks, leave status. Early evidence: completion rates shift meaningfully within one cleanup cycle.
2. Pick one business outcome your company already tracks and design learning measurement against it. Early evidence: you can explain in one sentence how a program moves a metric your CFO already cares about.
3. Use AI as your BI expert, not your L&D replacement. Tools like Databricks let you build dashboards iteratively by talking to AI — checking against your manual numbers until you trust the output. Early evidence: one self-built dashboard replaces one Crystal Reports request.