This is still a wip and results might not be fully accurate yet. Right now I’m just passing the user’s queries to Gemini / claude. And then querying the tmdb api with the results.

Next step is to embed a tmdb dataset and perform rag on it.