determining sections using embeddings

There are only so many sections a cv can have. We can have a description and title of them and compare the headers of the cv sections to them and pick the most semantically similar.

I am not sure if this is going to work but it is easy to try and manually validate.

No files, tests, or entry points are named. Start by locating the CV section parsing flow and checking how section headers are currently handled; compare headers with predefined section titles and descriptions using embeddings, then manually validate the resulting section assignments.

Written by the indexing model from the issue text.