GridGuard GridGuard is a Streamlit web app for short‑term electricity demand forecasting. It’s designed to help understand grid stress risks under different conditions.

What it does Predicts electricity demand 1, 3, 6, or 24 hours ahead Supports multiple regions (Austin, Dallas, Houston) Flags potentially risky or unrealistic inputs Shows simple confidence ranges and risk explanations Includes historical model validation with adjustable time windows

Why it matters Electricity grids are sensitive to spikes in demand, weather extremes, and timing. GridGuard helps visualize how demand may change and when the grid could be stressed.

How it works Uses trained machine learning models per region and forecast horizon Applies basic feature engineering (time of day, seasonality, temperature) Runs entirely in a single app.py Streamlit app

Tech stack Python Streamlit Pandas / NumPy Matplotlib Scikit‑learn (via joblib models)

Run locally pip install -r requirements.txt streamlit run app.py

Notes This project is for educational and analytical purposes, not real‑time grid operations.