Global
Anomaly Detection on supply chain
Anomaly Detection Model to Predict Group of Anomalous Products on Backorders
Overview

Anomaly Detection model to predict group of anomalous products on backorders

Background

For the supply chain area, BIP developed an anomaly detection model to predict groups of anomalies in the quantity of backorders for different products on multiple factories. The problem is approached with unsupervised learning techniques, leveraging Isolation Forest model.

Service & Capabilities

Automotive

Technology

Python
Amazon
Results

Results Achieved:

The solution is developed on AWS Cloud, using Step Functions to perform ingestion, anomaly computation and results saving. Every night, the model predicts the anomalous products saving the output in an AWS S3 buckets connected to a Qlik dashboard that allows the business to view and provide feedback on the anomalies. Moreover, the model identifies common characteristic that mostly represents anomaly group of products.​

Factories were integrated in the model​

14

Correct anomalies by business feedback​

~85%​

Ready to collaborate and create value?

Let’s get in touch and see a new vision of the future!

Enter your details, one of our experts will get in touch soon