RAG-Based Bitcoin Analysis Chatbot
GenAI
NLP
Docker
Finance
A real-time crypto analytics bot using Ollama, LLaMA3, and Docker.
Overview
This project integrates Large Language Models (LLMs) with real-time financial data to provide actionable crypto insights. [cite_start]Unlike standard chatbots, this system utilizes a Retrieval-Augmented Generation (RAG) framework to ground answers in live market data[cite: 30].
Technical Architecture
- [cite_start]Model: LLaMA3 served via Ollama[cite: 30].
- [cite_start]Vector Database: FAISS for vector retrieval using nomic-embed-text embeddings[cite: 32].
- [cite_start]Data Pipeline: Aggregates real-time prices (CoinGecko API) and news (NewsAPI)[cite: 31].
- [cite_start]Deployment: Fully containerized Streamlit application using Docker[cite: 33].
Key Features
- [cite_start]Sentiment Analysis: Applies NLTK to financial news to gauge market sentiment[cite: 31].
- [cite_start]Technical Indicators: Auto-calculates RSI, MACD, and Bollinger Bands[cite: 31].
- [cite_start]Forecasting: Integrated TensorFlow/Keras LSTM for short-term price prediction[cite: 32].
(Note: Add code snippets or GitHub links here)