Skip to content
Case study · Retail and CPG · Data and analytics

Supply chain analytics on Google Cloud for a leading retail firm

A leading retail firm connected SAP HANA to an analytics platform on Google Cloud, with machine learning demand forecasts at SKU and store level driving inventory decisions.

Industry
Retail
Capability
Data, analytics and ML
Platform
Google Cloud, SAP HANA
Related POD
AI POD

Context

The client is a leading retail firm. It engaged VerticalServe to design and implement a supply chain analytics solution on Google Cloud Platform (GCP) to improve demand forecasting and inventory management.

The problem

The firm had difficulty forecasting demand accurately and managing inventory levels. That made supply chain operations inefficient and led to higher costs, stockouts and lower customer satisfaction.

What was built

The solution has four components.

Flow: SAP HANA, historical sales and seasonal and external factors feed the Cortex data platform on Google Cloud, where demand forecast models and inventory algorithms produce SKU and store forecasts and inventory levels. INPUTS SAP HANA Historical sales Seasonal and external factors GOOGLE CLOUD Cortex data platform Demand forecast ML models Inventory algorithms DECISIONS SKU and store forecasts Optimized inventory levels

Flow: SAP HANA, historical sales and seasonal and external factors feed the Cortex data platform on Google Cloud, where demand forecast models and inventory algorithms produce SKU and store forecasts and inventory levels.

  • Cortex: a data analytics platform on GCP that aggregates and analyses supply chain data from various sources, including SAP HANA.
  • SAP HANA integration: the firm’s SAP HANA system is integrated so the solution works from real-time supply chain data.
  • Demand forecasting: machine learning models on GCP forecast demand at SKU and store levels from historical sales data, seasonal trends and external factors.
  • Inventory management: inventory algorithms set inventory levels from demand forecasts, lead times and replenishment cycles.

Stack

Google Cloud Cortex SAP HANA Machine learning models

How it was delivered

VerticalServe began by implementing Cortex on GCP to aggregate and analyse the firm’s supply chain data, and integrated the SAP HANA system as the real-time source.

Next the team developed and deployed the demand forecasting models on GCP.

The inventory management algorithms came last, since they depend on the forecasts. They combine forecasts with lead times and replenishment cycles to limit both stockouts and excess stock.

The result

  • More accurate demand forecasts, leading to better-informed supply chain decisions.
  • Optimized inventory levels, with fewer stockouts, less excess stock and lower associated costs.
  • Data-driven decisions from real-time supply chain data, through the SAP HANA integration and Cortex.
  • Lower costs from stockouts, excess inventory and inefficient replenishment cycles.

The original write-up reports these outcomes without figures.

What it means for a similar team

  • Connect the system of record first. Forecasts here were only as current as the SAP HANA data behind them.
  • Forecast at the level where stock decisions are made. In retail that is the SKU and the store.
  • A forecast changes nothing on its own. The inventory algorithms that act on it are where the cost reduction came from.

This case study was first published on our Medium publication. Original article

Engage

Start a similar POD

Tell us the outcome you want and your constraints. We will reply with a discovery proposal.

Talk to us

Working on something similar?

An Innovation POD is a dedicated team that builds inside your environment. Tell us what you need to ship.

Start a POD