Neo4j Morpheus

Neo4j Morpheus: Weave Tabular & Graph Data Together in Spark

Business data is on the move toward data lakes.

Connecting and harvesting the value of these data lakes is essential for digital transformation. Yet – according to Gartner – 90% of data lakes may become useless by the end of 2018 as data analysts are overwhelmed with trying to sift through information assets captured for uncertain uses.

Enter: Neo4j Morpheus.

With Morpheus, you‘ll get more out of your data lakes and streamline workflows between graph and relational worlds, allowing you to weave together tables and graph data in Spark for in-memory analysis.

Morpheus serves as the key data integration layer between the Neo4j Graph Platform and Apache Spark. In addition, it features Cypher for Apache Spark, extending Cypher‘s reach to Spark and its full range of tools and analytics.

Check out the Morpheus User Guide

How Morpheus integrates into the Neo4j Graph Platform

The Challenge of Enterprise Data Integration

Morpheus data integration

It‘s no easy task to wrangle data created for one purpose into another — let alone combine multiple models. Thus, the core challenge of enterprise data integration.

Neo4j Morpheus overcomes this challenge by creating a graph lens that quickly reveals connections and fresh data discoveries between disparate data types and sources.

Using Morpheus, your data scientists can analyze multiple graphs at once as well as mix graph and tabular data — without the burden of copying or moving data — and then save those integrated supersets or subsets in HDFS, other file systems or Neo4j. This enables graph analysis across different data types while simultaneously keeping data in situ, elevating what‘s possible for complicated integration challenges.

How Neo4j Morpheus Delivers More Value from Enterprise Data Lakes

Morpheus delivers value

To extract more value from your data lakes, Morpheus uses Cypher for Apache Spark with advanced aggregation, filtering and pattern matching. This means that your team uses a more intuitive graph model and the declarative Cypher query language to reveal connections across your data lake — instead of struggling with complicated JOINs.

Here are a few ways Morpheus delivers more business value from your enterprise data lakes:

  • Multiple named graphs allow your data scientists to quickly discover relationships among different data sources.
  • Scaling summarizations (i.e., multiple views on a single graph) give you the power to quickly switch between perspectives, such as comparing different end-of-day results or day vs. month results, all without creating new graphs.
  • Combine SparkSQL and Cypher statements — using whichever matches the comfort of your analysts and the context of the query.
  • Progressive graph analysis adds information and value at each step by using an output graph as the input graph for your next analysis when using composable queries.
  • Lift graph data from multiple sources and formats , perform your transformations and then store results to other uses.
  • Write Morpheus results back to HDFS or other file systems , export data to tabular files or store/merge graphs to Neo4j.
  • Migrate Hadoop data through Morpheus to Neo4j for transactions and structural analysis.

Get Early Access to Neo4j Morpheus

Morpheus data integration

Neo4j Morpheus brings features from the most popular graph platform to one of the most popular distributed compute platforms: Apache Spark. Morpheus transforms the value of your data lake because you‘ll immediately see what‘s been hiding in plain sight.

General availability of Neo4j Morpheus is planned for autumn 2018.

We are accepting a limited number of customers in our early adopter program. If you would like to participate and have the basic environment below, please contact your account representative.

Apply now for the Morpheus Early Access Program (EAP) with software available in early November.

Sign Up

Contact your account representative for more information.

Contact Us

Basic Environment Elements

Required Software:
  • Apache Spark, preferably 2.2 or later
  • Scala v2.11
  • Neo4j, preferably v3.4 or later
Some mix of storage libraries:
  • Parquet
  • Hive
  • Neo4j
  • Spark familiarity
  • Use of Spark in a development environment
  • If using HDFS or Hive, have an existing development environment
  • Use of Scala as an application development language