I am actually working for a research project in the medical field (CDSS) in partnership with the industry. RDFox has been chosen and I use it all the time. I am always impressed by its power and speed of execution. For example, a request time of several minutes with other SPARQL engines is often solved in less than one second with RDFox !

RDFox is a "lean and mean" semantic graph database, Oxford Semantic Technologies clearly understands what the key priorities are for anyone who wants to deploy a database in production: it has to be fast, reliable and predictable. Many of their competitors focus on adding more and more (often proprietary) features forgetting (a bit) about these three key priorities. RDFox has a radically different design than other semantic graph databases in the sense that it is an "in-memory" database (with full ACID transaction support) which means that it is screaming fast (many LUBM queries are more than a 1000 times faster than the nearest competitor) but it also means that it really needs a lot of memory for large datasets. Fortunately, if you'd be building an Enterprise Knowledge Graph (EKG) according to the 10 principles of the EKGF (see https://ekgf.org/principles) you would not need to have all your datasets in just one database instance and could freely scale horizontally. RDFox would allow you to support real-time use cases like pre-trade risk calculation or other advanced use cases using many complex datasets.

High-performance SPARQL querying and Datalog reasoning even with large datasets. In-store SHACL engine that is easy to use. We integrate our software platform and develop real-time recommender use cases, which are made possible because of the fast incremental reasoning.