28%Inventory Cost Down
18%Delivery Time Saved
97%Forecast Accuracy
14Sources Unified

The Challenge

A major manufacturing company operated across 6 production sites with 14 separate data sources — ERP systems, supplier portals, logistics APIs, and warehouse management tools — none of which communicated with each other. Supply chain planners spent 12 hours weekly manually pulling and reconciling data in Excel to produce demand forecasts that were already 48 hours stale by the time leadership reviewed them.

Inventory levels were poorly optimised: some SKUs were chronically overstocked while others caused production stoppages due to stockouts. The CFO challenged the operations team to reduce carrying costs and improve delivery performance simultaneously.

Our Approach

1

Data Landscape Discovery

Catalogued all 14 data sources: 3 SAP ERP instances, 4 external supplier APIs, 3 logistics providers, 2 WMS systems, and 2 Excel-based planning tools. Assessed data quality and transformation complexity for each.

2

Snowflake Lakehouse Architecture

Designed a three-layer Snowflake architecture: raw ingestion, cleaned staging, and business-ready semantic layer. Configured secure data sharing for external supplier portal integration.

3

dbt Transformation Pipeline

Built 87 dbt models transforming raw operational data into clean, documented business metrics including inventory turnover, supplier lead time variance, and demand volatility scores.

4

Demand Forecasting Model

Developed a Python-based statistical forecasting model (SARIMA + external seasonality factors) trained on 3 years of demand data. Integrated with Snowflake via a scheduled stored procedure orchestrated by Airflow.

5

Power BI Dashboards

Delivered 6 Power BI dashboards providing live visibility into inventory levels, supplier performance, logistics KPIs, and procurement recommendations for the planning, operations, and executive teams.

The Results

The platform gave the supply chain team a single source of truth for the first time in the company's history. Planners now spend less time collecting data and more time acting on it.

  • Inventory carrying costs reduced by 28% through optimised reorder points
  • Average delivery time improved by 18% due to better supplier lead-time visibility
  • Demand forecast accuracy reached 97% — up from 61% with the legacy Excel process
  • Planning team saved 48 person-hours per week previously spent on data consolidation
  • Stockout incidents reduced by 63% in the 6 months post-launch

“For the first time, our planners, logistics team, and finance department are all looking at the same numbers at the same time. The Snowflake platform RevOps built is now central to every supply chain decision we make.”

— VP Supply Chain & Operations, Manufacturing Enterprise

Ready for Results Like These?

Let’s discuss how we can solve your unique challenges and deliver measurable impact.

Start a Conversation View All Case Studies