Bitcoin Price Analysis with Ollama

Built from March to May 2025 at the University of Maryland, this project brings live Binance price data into a Python analysis pipeline.

I used pandas to calculate moving averages, volatility, and anomalies. A local Mistral 7B model, called through Ollama-Python, produced natural-language summaries and short-term forecasts. Streamlit provided the dashboard, with Docker for deployment.

The project explored how numerical time-series analysis and local language models could work together in an interactive application.