Scenario (Background)

A mid-sized service organization with a mature Lean management system saw demand rise without added staffing. Lean practices were strong (daily huddles, visual management, CI), but work flowed across multiple digital systems with frequent hand-offs. Performance data existed, but it was fragmented and reviewed after-the-fact. Managers estimated they spent 10–15% of their time manually preparing data instead of managing flow and coaching teams.

Challenge (Lean Problem)

The value stream review showed recurring waste:

• Waiting at hand-offs (queues hidden until service levels were breached)
• Rework/defects (25–30% of cases required rework due to missing/incorrect information upstream)
• Overprocessing (duplicated reporting and inconsistent “versions of the truth”)
• Reactive problem-solving (issues identified retrospectively, slowing PDCA)

Solution (AI-Enabled Lean Approach)

AI was introduced in three Lean-guided ways:

• Process mining to reveal actual flow, hidden queues, and rework loops
• Predictive models to anticipate bottlenecks and smooth workload
• AI-supported CI tools to summarize performance signals and draft initial A3 problem statements, while teams retained ownership of analysis and countermeasures

Key Learnings

• Lean didn’t need replacing, rather visibility and speed of insight did
• Earlier detection improved flow and strengthened PDCA
• The best results came from targeted AI use, anchored to Lean principles and governance

Discussion Questions

1. Where do hand-offs create invisible queues in our processes?
2. What rework loops do we accept as “normal” and why?
3. What data preparation could be automated so leaders can coach more?
4. Where could earlier signals prevent missed targets?