How krish27naa/telecom-churn-prediction works
This is a Streamlit customer-churn risk application. An operator enters customer attributes, submits them for XGBoost inference, and receives a probability, risk category, retention priority, tailored recommendation, and customer summary. The runtime is concentrated in App.py; the persisted XGBoost model is the only model artifact the sampled code loads. The tree also contains training and reporting artifacts, but no README or source evidence connects them to this runtime, so they are not shown as active workflow components.
Main entry points: App.py Churn Scoring · xgboost.pkl XGBoost Model
Components 8
Customer Input
- Customer Form
AI_Project/AI_Project/App.py - Customer Record
AI_Project/AI_Project/App.py
Churn Assessment
- XGBoost Model
AI_Project/AI_Project/outputs/models/xgboost.pkl - Churn Scoring
AI_Project/AI_Project/App.py - Risk Classification
AI_Project/AI_Project/App.py
Retention Results
- Retention Guidance
AI_Project/AI_Project/App.py - Results Display
AI_Project/AI_Project/App.py
Other components
- Operator
Connections 10
- Operator → Customer Form: enters details
AI_Project/AI_Project/App.py - Customer Form → Customer Record: builds record
AI_Project/AI_Project/App.py - Churn Scoring → XGBoost Model: loads model
AI_Project/AI_Project/App.py - Churn Scoring → XGBoost Model: requests probabilities
AI_Project/AI_Project/App.py - Churn Scoring → Customer Record: scores record
AI_Project/AI_Project/App.py - Churn Scoring → Risk Classification: classifies score
AI_Project/AI_Project/App.py - Risk Classification → Retention Guidance: selects action
AI_Project/AI_Project/App.py - Risk Classification → Results Display: provides risk
AI_Project/AI_Project/App.py - Retention Guidance → Results Display: provides recommendation
AI_Project/AI_Project/App.py - Operator → Churn Scoring: submits prediction
AI_Project/AI_Project/App.py