How mrinalgaurkhede/mlops_telecom_customerchurnpredictionmodel works
This is a telecom customer-churn ML template organized around data ingestion and preprocessing, with documented training, evaluation, prediction, CLI, and web-serving capabilities. The sampled runtime code verifies ingestion and preprocessing; README claims describe the broader workflow, but the listed model-stage files are empty and the documented API source is absent from the tree, so those relationships are shown only where documented.
Main entry points: main.py Pipeline CLI · model_evaluation.py Model evaluation · src REST API · processed Train and test data
Components 16
Data preparation
- Raw churn data
data/raw - Data ingestion
src/data_ingestion.py - Ingested data
data/interim - Data preprocessing
src/data_preprocessing.py - Train and test data
data/processed
Model workflow
- Model training
src/model_building.py - Versioned models
models - Model evaluation
src/model_evaluation.py - Evaluation reports
reports - Model prediction
src/model_prediction.py
User interfaces
- Pipeline CLI
main.py - REST API
src - Web console
frontend/index.html
Configuration
- Pipeline configuration
config.yaml
Other components
- CLI user
- Browser
Connections 22
- CLI user → Pipeline CLI: runs stages
README.md - Pipeline CLI → Data ingestion: calls
main.py - Data ingestion → Raw churn data: reads CSV
src/data_ingestion.py - Data ingestion → Ingested data: writes data
README.md - Pipeline CLI → Data preprocessing: calls
main.py - Data preprocessing → Ingested data: reads data
README.md - Data preprocessing → Train and test data: writes splits
README.md - Pipeline CLI → Model training: runs stage
README.md - Model training → Train and test data: reads splits
README.md - Model training → Versioned models: saves model
README.md - Pipeline CLI → Model evaluation: runs stage
README.md - Model evaluation → Train and test data: reads splits
README.md - Model evaluation → Versioned models: loads model
README.md - Model evaluation → Evaluation reports: writes metrics
README.md - Pipeline CLI → Model prediction: runs stage
README.md - Model prediction → Versioned models: loads model
README.md - CLI user → REST API: sends requests
README.md - Browser → Web console: uses
README.md - Web console → REST API: sends HTTP
README.md - REST API → Versioned models: loads cached model
README.md - REST API → Train and test data: reads features
README.md - Pipeline CLI → Pipeline configuration: uses settings
README.md