AI powered metadata scraping

[Janne Kario]

I've accumulated a collection of TVHeadend recordings over the years (terabytes, hundreds of movies and series) and mounted the recordings folder to my Jellyfin. I would like to have all the metadata for the movies in place. The problem is that I have to do it all manually. The filenames come directly from the Finnish EPG. For example "Elokuva: Alien: Covenant (16).ts". Jellyfin does not automatically identify the file. I can manually identify the file by selecting the Identify option, doing a search using the movie name and selecting the correct movie from the search result. However, I would not like to do this 1000 times by hand. I gave this task to chatgpt instead. I gave it the filename and said I wanted the IMDB id of the movie and it could produce it with almost 100% accuracy. Is there a way to automate this using some kind of API?

[TheDreadPirate]

You can try using Tiny Media Manager. It may still require manual work to correctly identify your files, but it should automate a lot of the more tedious tasks of finding IDs and renaming/reorganizing files.

Also, you should use TMDB IDs where possible.

[gnattu]

> Is there a way to automate this using some kind of API?

Yes. You can do it with a metadata provider plugin if you want to implement it.

[KrisMarshall]

For automation, APIs like OMDb or TMDb are great—they let you send a movie title and get back metadata, including IMDb IDs. Once you script that, you could batch process your files. Saw a neat write-up on https://openeverything.us/ about AI tools streamlining media libraries—it sparked some ideas to automate my own old photo tagging project.