The LLM writes de Query. Queries are better over the knowledge graphs than over a relational database.

If you can do data modeling in sequel then you can do ontology.

LLMs help us build the ontologies and catalogues

Tim Berber’s Lee was talking about trust. It is the key here.

Can the LLMs answer using your knowledge graphs than?

Noooo, it should be building queries!!!

Add the trustfulness sources within the connectors, and system of records!

This queries can be auditable!

The holy grail are the explanations from your data!