InventoryIQ: Retail Stock Analytics

The Problem

Retail operations managing multiple warehouses face a double-edged risk: running out of fast-moving, high-margin SKUs (stockouts costing revenue and customer loyalty) while tying up hundreds of thousands of dollars in dead capital for slow-moving products. Basic ERP reports provide point-in-time snapshots, but lack predictive safety stock formulas accounting for supplier lead-time variability.

I built InventoryIQ to give supply chain leaders full control: clean SQL ingestion, Star Schema data modeling, and automated DAX measures that calculate dynamic Reorder Points (ROP) and Days of Inventory Outstanding (DIO).

Data Architecture & DAX Modeling

The solution is built on clean dimensional principles:

1. Star Schema Model: Isolated central `Fact_InventoryMovements` and `Fact_Sales` tables surrounded by `Dim_Product`, `Dim_Location`, `Dim_Supplier`, and `Dim_Date` dimensions.
2. Optimized DAX Measures: Utilizes variables (`VAR`) and filter caching to compute dynamic ABC classifications (Class A: 80% revenue, Class B: 15%, Class C: 5%) with zero performance degradation across 100k+ SKUs.
3. Automated Reorder Trigger Logic: Compares real-time available stock against `(Average Daily Usage * Lead Time) + Safety Stock`.

The Stack

Business Impact

• -34% Reduction in Stockouts: Automated alerts enabled replenishment before inventory dropped below safety thresholds.
• $140,000 Working Capital Unlocked: Identified and liquidated dead Class C inventory.
• Sub-Second Report Filtering: 100% responsive across executive and warehouse tablet views.