Update README.md

README.md

@@ -12,3 +12,25 @@ short_description: 'Fraud detection system that detect fraud and explain it '
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# GuardianAI: Fraud Detection & Explainability
+
+This project uses a **Random Forest Classifier** to detect fraudulent transactions and **Google Gemini** to provide natural language explanations for the flags.
+
+### How to Setup:
+1. Create a Space on Hugging Face (Gradio).
+2. Upload all files.
+3. **API Key:** Go to Settings -> Variables and Secrets -> Add a secret named `GEMINI_API_KEY` with your key from Google AI Studio.
+4. Run `main.py` locally once to generate the `.joblib` file, or include it in your upload.
+
+### Features:
+- Real-time risk scoring.
+- GenAI-powered explanation for transparency (Explainable AI).
+- MIT Licensed.
+  # GuardianAI v2: Velocity Attack Monitor
+
+A fraud detection prototype for 2026 that detects **high-frequency attacks**.
+
+### New Features:
+- **Velocity Tracking:** Detects "Card Testing" or "Bot Attacks" where many transactions occur in a short window.
+- **Explainable AI:** Uses **Gemini 2.5 Flash** to provide human-readable reasons for why a card was blocked.
+