How rakshitv-12/blinkit-sales-analysis works
This repository analyzes Blinkit grocery sales and exposes a Streamlit sales-prediction app, alongside notebook-based EDA and model development and a documented BI dashboard. The main user flow is to enter product and outlet attributes, obtain a model prediction, and compare it with historical sales benchmarks and category visualizations. The app’s dataset loading and prediction behavior are visible in the excerpt; notebook and dashboard details are based on README documentation, so their internal flows are not expanded.
Main entry points: app.py Streamlit interface · Tableau BlinkIT Grocery Project U16955293080 (4).xlsx Sales dataset · sales_prediction_model.joblib Ridge pipeline artifact · Model_Prediction.ipynb Model training and evaluation
Components 10
Prediction app
- Streamlit interface
app.py - Sales dataset
data/Tableau BlinkIT Grocery Project U16955293080 (4).xlsx - Product and outlet inputs
app.py - Sales prediction
app.py - Benchmarks and charts
app.py
Sales analysis
- Exploratory analysis
notebooks/blinkit_sales_eda.ipynb
Model development
- Model training and evaluation
notebooks/Model_Prediction.ipynb - Ridge pipeline artifact
notebooks/models/sales_prediction_model.joblib
BI dashboard
- BI dashboard
dashboard
Other components
- App user
Connections 10
- App user → Streamlit interface: uses
README.md - Streamlit interface → Sales dataset: loads
app.py - Streamlit interface → Product and outlet inputs: collects
app.py - Product and outlet inputs → Sales prediction: predicts from
README.md - Streamlit interface → Ridge pipeline artifact: loads
app.py - Streamlit interface → Sales prediction: runs inference
app.py - Streamlit interface → Benchmarks and charts: renders
app.py - Sales dataset → Benchmarks and charts: supplies benchmarks
app.py - Sales dataset → Exploratory analysis: analyzed in
README.md - Model training and evaluation → Ridge pipeline artifact: serializes
README.md