AI-Powered Content Update Checker

This is an overview of a project I worked up for a Hackathon. Whenever a help center article is updated on the website on WordPress, this tool uses an embedding LLM model to take the contents of the article and checks it against all existing content. It finds the most relevant content to the article, and makes recommendatons of what sections might need to be updated. It uses semantic search and vector databases to help quickly find relevant content. Once the updates are identified, the tool sends a message through Slack indicating what content needs updating, what sections, and a reasoning behind the decision to include it in the list.