Neo4j Inc., the startup behind the widely used database of the same name, today released a new version of the system that will enable companies to take advantage of its specialized architecture in more ways.

Neo4j is a graph store, a kind of database specifically designed to keep track of the relationships between records. That makes the system adept at powering use cases such as fraud detection that require connecting the dots among many individual details.

Neo4j 3.4 brings several features designed to help users identify patterns more easily, starting with a new visualization tool called Bloom. It can take a search query and automatically display the data connections that come back as color-coded clusters (pictured). The tool is touted as simple enough for everyday business workers to use.

Bloom is complemented by a set of new search functions that are becoming available as part of Neo4j’s Cypher query language. The system now provides the ability to filter data based on the time and date logs associated with entities, as well as how they’re positioned in their three-dimensional environments.

Neo4j said the enhancement enables a range of new use cases. A logistics company, for example, might correlate the time drivers take to reach their destinations with other information collected from routes to find ways of improving delivery times. And using the new geospatial data support, a farm operator could account for factors such as elevation when making crop management decisions.

The new analytics features are joined by a host of architectural improvements. Among them is support for running multiple graph processing environments on a deployment and and up to a 50 percent speed boost for data loading and writes. The company has also added several new management features, most notably the ability to roll out updates one server at the time, which removes the need to take the entire deployment offline for maintenance.

Image: Neo4j