Very interesting development! Some kind of Prolog on relational database, even the syntax is similar. I can easily imaging use cases for this, but I wonder whether it can be really performant on big enough databases.

For example, triple datastores for RDF + rules engines looks more mature and advanced and probably SPARQL provides all the same possibilities (plus possibility to connect to several databases at the same time).

What is the sweet spot of Logica? Is it when put on top of relational db posed to replace graph databases? (I doubt)

What powers googles knowledge graphs then?