Retail Demand Forecasting Tool

Advanced analytics tool built to predict weekly product demand across stores using Azure SQL, Python ML, and a small C# API gateway.

Features

• Train baseline Random Forest model (scikit-learn) and optional TensorFlow DNN.
• FastAPI-based inference service (Python).
• C# (.NET 8) minimal API gateway that proxies inference.
• Dockerized services and sample GitHub Actions CI/CD for pushing container images and deploying to Azure Web Apps.
• SQL Server / Azure SQL schema + seed data.

Quickstart (local)

1. Clone the repo.
2. Create a .env from .env.example and fill values.
3. Populate the database: run sql/01_schema.sql then sql/02_seed.sql on your SQL Server/Azure SQL.
4. Train model:

   cd ml
   pip install -r requirements.txt
   python train_sklearn.py