Stop Stocking for Parts That Never Fail: Purpose-Trained AI Agents for MRO Inventory and a Reality Check from Manufacturing Customer
Tuesday, December 8, 2026, – (America/New_York)
Level 3: Calusa 5
Most storerooms protect the wrong parts. Min/max rules written years ago keep planners overstocked on items that never fail and exposed on the spares that stop production and the criticality list that was supposed to fix it lives in a spreadsheet nobody has updated since the last turnaround.
This session shows what purpose-trained AI agents do about that, specifically: score every spare against the asset hierarchy and failure impact instead of last year’s usage; sense demand from real maintenance signals: PM plans, open work orders, the shutdown calendar rather than consumption history alone; confirm the required spares are on hand before a work order is released; and find the part sitting idle at a sister plant before a PO is cut. We will be equally specific about where the agent stops and the planner decides, and about what these systems cannot do.
You will leave with a five-question scorecard to run against your own storeroom on Monday, the criticality questions to put to any vendor, and an honest list of the limits. Honest about the limits, specific about the gains.
What you’ll take away
- The five questions that tell you in an hour whether your storeroom is protecting the right parts
- How agent-scored criticality differs from the spreadsheet and why the difference is measured in downtime
- A practitioner’s account of the first weeks on real data: the findings, the objections, the decisions
- Where humans must stay in the loop, and what to refuse from any vendor who says otherwise
Knowledge Level: All-Inclusive
Speakers
-
Director, Financial Planning & Analysis, Smurfit Westrock -
Chief Executive Officer, Verdantis