This week, Neo Technology Inc. announced they had raised $36million in a series D funding round. Neo Tech will now use these funds to enhance their leading product, Neo4j.
For those that have been away for some time, Neo4j is the leading Graph database in 2016, and is continuing to make a rapid rise, in this self created "graph database" category. Neo4j is best described as an open source ACID compliant graph database, with Enterprise and Startup license options available, with downloads now reaching 2.5 million.
Neo4J uses an emerging standard query language, Cypher, which makes it very easy to express rich graph traversal queries, which can be very effective at supporting analytics solutions. It is open source (GNU GPLv3), has a well established and large user base and has a very active community and conference/meetup circuit.
In CEO Emil Eifrem's blog post, there is some description as to how he sees the database market evolving towards 2020, broken into 3 market segments (relational, what we know today as nosql, and more niche markets), and with most players having open source alternatives of their full versions, and communities attached to each.
OverStory have been looking at graph databases in general and running trials on Neo4j for some time now. It's property graph lends itself to many uses, many of which are documented on Neo4j's site. What we especially like, is its ease of entry, and application potential. In the trials we have found that the property node graph model is especially well suited to exploring linked datasets and finding patterns or to discover unexpected connections. And Neo4j is demonstrating this with its success in areas such as forensics, recommendation algorithms, fraud detection, analytics and so on.
The rise of the graph category, and Cypher as a query language doesn't mean that RDF/SPARQL is going anywhere, as its an extremely powerful combination, however the proprietary nature of Cypher has made us watch closely what was happening in this space. With past academic articles debating the performance merits of property graphs vs RDF triple stores, and more recently Kurt Cagle suggesting that there will be growth in graph for some time to come, it's a technology we are fully embracing at OverStory.
We are very excited by the potential of this database and will be demonstrating use cases in the near future. In the meantime, if you want to find out what we could do for your business, please don't hesitate to get in touch.