This project is a fan-built website that presents media discussed on the Very Bad Wizards podcast.
The site is built using Next.js and styled with Chakra UI.
The media lists are generated by processing the podcast's RSS feed through the OpenAI API. The API analyzes the text to identify the media discussed in each episode. The extracted data is then stored in a database (MongoDB) for easy retrieval and display on the site.
As of now, there may be some errors in the data given the nature of the LLM, but the site updates daily to reflect the latest episodes and their associated media. The accuracy of the media extraction is expected to improve significantly upon gaining access to newer models of the API.
Stay tuned for more updates! Contributions are welcome!