Wikimedia Foundation is looking for a specialist in Graph databases and/or RDF stores in order to support Wikidata, the largest and most actively edited open Knowledge Graph in the world for a five month consultation contract.
Wikidata had more than 400,000 contributors and data from Wikidata is seen by more than a billion people every month. The SPARQL endpoint contains more than 12 billion triples and answers millions of queries per day. Wikidata is the most edited wiki in the world with 1.5 billion edits. You work in a completely open source environment and can openly share what you are doing. Check out our public dashboards about the query service:
https://grafana.wikimedia.org/d/000000489/wikidata-query-service?orgId=1&refresh=1m
We are a non-profit supporting an active global community. And we're having a really big challenge to crack! If this excites you, if you can and want to help Wikipedia - Join us!
And / or spread the word!
What's the motivation for moving away from Blazegraph? Have you hit some limits of scalability?
The company behind Blazegraph has been acquired by Amazon, and they turned the software into Amazon Neptune. Since then Blazegraph itself hasn't received anything near the previous care and love, and the bitrot is getting worse. So we are noticing the aches more and more - and it is a bit unclear whether it is scalability or bitrot or both. Since Wikidata was always one of the most demanding users of Blazegraph, we have been relying on the support of its development team to get where we are. It might be that it could scale further - but not without the support. Also, the Open Source version of Blazegraph doesn't support sharding, which means that partially scaling meant buying a bigger machine. This also has come to the point where it isn't feasible anymore.