Use of an LLM for extracting metadata from datasets

[marcel-vesely-kinit]

This task focuses on one of the main functionalities of the whole metadata filtering feature, that being the automated extraction of predefined set of metadata filters of datasets from JSON format. As an initial step, we need to implement this functionality on datasets asset type type only for now, and once we're satisfied with the implementation, we can further extend it to other asset types as well

BLOCKERS:

• This feature needs to be done first (at the very least for the dataset asset type) => Identify lists of metadata fields specific for each asset type for metadata filtering #21