TELECOMMUNICATIONS COMPANY
BUSINESS PROBLEM
This telecommunications company aimed to grow its customer base and increase profitability through customer segmentation analysis. However, two key challenges needed to be addressed before proceeding:
The necessary data for the analysis was dispersed across multiple cloud-based and on-premises applications, each utilizing different databases.
The data sources lacked standardized fields, making it difficult to match observations across the applications.
SCALESOLOGY IN ACTION
To address the first issue, a dynamically scalable and expandable data lake architecture was deployed in a modern DevOps environment using Azure Storage and Snowflake. This setup was designed to support current and future data consumption needs over the next 3–5 years. Data import and replication tools, developed in Python, refresh data daily from eight applications. Additionally, a CI/CD pipeline implemented through GitHub automatically tests and deploys changes to an Azure Function App.
The second issue was resolved using a combination of regular expressions (regexp) and fuzzy string matching algorithms written in SQL and Python to identify and match records across data sources. This effort also revealed several data quality issues within the systems that required resolution prior to conducting the analysis.
RESULT
The telecommunications company successfully developed and implemented a customer segmentation scoring model to enhance current customer engagement and identify promising, untapped business opportunities.
SERVICE REFERENCES
留言