What problem does it solve? Fill rate holds at target but days-of-inventory keeps climbing because safety stock sits in the wrong echelon relative to where demand variability now lives. This Skill re-segments demand variability by SKU-region-channel, re-solves safety stock per node using measured demand correlation, and prices re-positioning options on freed working capital versus service risk. ## Core Features & Use Cases - Multi-echelon re-cut: Computes safety stock per node with σ_DDLT and the correlation-adjusted pooling formula, deciding central vs forward-deployed placement per segment. - Priced options with a human gate: Builds three priced positions (forward-deploy, aggressive central pool, selective top-SKU), stages targets, and holds publish for human approval since publishing re-drives MRP. - Failure-to-recovery playbook: Covers mis-set safety stock, ignored demand correlation, freight eating the saving, variability noise churn, and destructive overwrites of live published targets. - Use Case: Days-of-inventory rose 41 to 49 over three months on a $210M base while fill rate held at 98%. The Skill re-segments store vs e-comm variability, thins smoothed store and DC buffers, pools volatile e-comm at a regional node, and frees roughly $8.5M in working capital at held service. ## Quick Start Ask the assistant to analyze where safety stock should sit across the network given that fill rate is on target but days-of-inventory has climbed for three straight months.
