I have a question about Neo4J viz Elasticsearch. Is anything lost if one uses ES and not Neo4J, functionally? Performance? Somewhat related, if someone has already invested in a different graph database, is there any support or planned support for something other than Neo4J?

Hi! We generally recommend using ES as the graph DB at this time. It simplifies deployments, does not require dealing with Neo4J's licensing, and we have not found it to be a significant performance drag on our deployments. Being not a native GraphDB, complicated deep level graph queries are almost certainly going to be slower on ES than on a native GraphDB, but most use cases do not require extremely high levels of hops. There was some initial work that was done to support DGraph as a graph backend, but it has not seen much battle-testing nor have we seen a particularly large demand for this from the community.