LLM Sentiment Analysis

Developed a Python web scraper to obtain news articles about a list of subjects, saving HTML and plain text versions of the articles in a PostgreSQL database.

Deployed Mixtral-8x7B large language model (LLM) for sentiment analysis of scraped articles using ollama. Wrote Python scripts to prompt the LLM and determine the sentiment of the set of articles. The results of the LLM analysis were stored in the same PostgreSQL databases.