Customer Segmentation – Unsupervised Learning – DATACREW.AI

Client:

An established Bank with 14 million Customers

Project Title

Building a Unified and Secure Data Foundation for a Leading Bank

Client

An established Bank with 14 million Customers

Problem

Data scattered across systems (transactions, demographics, activities).
Difficulty in unifying diverse data sources.
Ensuring data privacy, especially with personally identifiable information.
Struggling to derive actionable insights for strategic decisions.

Solution

Extracted data from multiple internal sources, using Spark, Hadoop, and ETL to store data in a centralized data lake.
Established a secure on-premise Data warehouse storage with a relational database.
Rigorous cleansing, de-duplication, and harmonization.
Engineered new features. Model development using K-means and hierarchical clustering.
Trained models, applied for segmentation and in-depth analysis.

Value Proposition

Tailored marketing for higher response rates.
Develop personalized products based on customer segments.
Higher customer satisfaction and retention.
Optimize resource allocation, hence improving efficiency and reducing costs.

Data integration, migration, master data management, data lakes & warehouses, IoT data platforms.