This project builds a serverless data analytics pipeline on AWS to analyze customer churn using the Telco Customer Churn dataset from Kaggle. The system ingests raw data, processes it into an analytics-friendly format, and enables SQL queries and dashboards to gain insights.
- Amazon S3 – stores raw CSV input and processed Parquet output
- AWS Glue – crawler to catalog schemas and optional ETL job for Parquet conversion
- Amazon Athena – SQL queries and views for churn KPIs
- Amazon QuickSight – dashboards for visualizing churn insights
- Amazon EventBridge (optional) – schedule Glue ETL jobs
- Terraform – Infrastructure as Code for reproducible setup
https://www.loom.com/share/410b4ecedb664fc9bb0976f37f07032e?sid=b1e67724-c986-4a87-9274-fd314052f2f2
For a deeper dive into the challenges faced and design decisions, see the blog post here: https://dev.to/ajithmanmu/how-i-built-a-serverless-data-analytics-pipeline-for-customer-churn-with-s3-glue-athena-and-bfk