Real-Time Crypto Sentiment & Volatility Engine (FinBERT)

This project utilizes FinBERT (Financial NLP Model) to perform sentiment analysis on cryptocurrency news and correlates it with real-time price volatility. It unifies market sentiment and technical analysis (SMA) into a single, interactive dashboard for investors.

Features

• AI-Powered Analysis: Classifies news as Positive, Negative, or Neutral using the FinBERT model.
• Live Data Stream: Real-time prices via Kraken and breaking news via RSS feeds.
• Data Persistence (SQLite): Analyzed news is stored in a local database, ensuring no data loss on refresh.
• Technical & Fundamental: Visualizes news markers directly on price charts alongside SMA (20/50) indicators.
• Interactive Dashboard: A modern UI built with Streamlit and Plotly.

Streamlit App Link

https://crypto-finbert-engine-mctlujtlgmfhepdmmdes5m.streamlit.app/

Installation

1. Clone the repository:

   git clone [https://github.com/damraka/crypto-finbert-engine.git](https://github.com/damraka/crypto-finbert-engine.git)
   cd crypto-finbert-engine

2. Install dependencies:

   pip install -r requirements.txt

3. Run the application:

   streamlit run app.py

📸 Screenshots

The application includes Price Action Charts, Sentiment Distribution (Pie Chart), and Word Clouds.

---

Developed by Damra Kaan Aglamaz