Drug and Vaccine Discovery: Knowledge Graph + Apache Spark

[@mine4me403]

cool, made GSK another 2 billion, congrats!

[@tuba-inxs]

I wonder if Apache Hudi could be replaced 1:1 in your implementation with Amazon's Neptune RDF Triplestore (I don't work for Amazon, and I think it might be interesting, since it wouldn't require managing Hudi).

[@johnhunter1252]

Neptune does not fit our use case. The main KG is too large for Neptune to support, and we needed point in time and incremental querying.