This Week in Neo4j – Behind the Trillion Graph, Mocking Neo4j in JavaScript, Bloom Tricks, Reactive Driver, Cloud Integrations, Data Science Talks, Neo4j Review, AuraDB Quickstart Videos

Hello, everyone!

Even with the start of the summer and NODES behind us, there is no pause of activity in the Neo4j community. This week we’ve seen a lot of articles, projects, and initiatives popping up.

This week’s featured member Yisroel Yakovson just published a mocking library for the Neo4j JavaScript driver that you should check out.

One of the big announcements was the Trillion Relationship Graph demonstration – in Chris Gioran’s article we can see what happened behind the scenes. Then we look at some advanced Bloom topics, with Dan Flavin magic search queries and Angeliki’s explanation of the reactive JavaScript driver usage for large data volumes.

For those of you who’d rather watch videos, we have all the talk recordings of the NODES data science track with knowledge graphs, link prediction, and dream analysis, followed by Ashley Faith’s Neo4j review video (featuring Dave Voutila) as well as a new set of AuraDB intro videos that you can share with your colleagues and friends. David Allen was a guest on the GraphStuff.FM podcast, discussing integrations with Apache Spark and Kafka as well as Bi Analytics and GraphQL APIs with Will Lyon.

And prolific talk commentator par excellence, Will Lyon shares the content of his NODES talk in a Twitter thread.


This week’s featured community member is Yisroel Yakovson.

Yisroel Yakovson - This Week’s Featured Community Member

Yisroel Yakovson – This Week’s Featured Community Member

Yisroel lives in Jerusalem, Israel and is CEO of NoStack, an online service allowing front-end developers to manage a full virtual stack. They integrate with Neo4j and are using GraphQL with GRANDstack. In that function, he provided a lot of feedback for the GraphQL integrations we built at Neo4j.

Yisroel has been an active member of the Neo4j community for quite some time. He published his first articles in 2018 explaining the power of “Full Graph Servers” using GraphQL and Neo4j. This week he published an article about “Mocking Neo4j calls in Node Applications.” He explains how unit testing with Neo4j was not as flexible and easy as he wanted and how he went ahead building neo-forgery a testing library to mock calls of the official Neo4j JavaScript driver. Please try it out and let him know how well it worked for you.

Yisroel, thanks a lot for all your work at NoStack, content, and projects in the Neo4j community.

Behind the Scenes of the Trillion Relationship Graph

trillion graph

One of the big highlights of NODES was the demo of the Trillion relationship graph in Emil’s keynote. And not just because of the coffee-mug-meme. The engineering team put together that scale demonstration within three weeks, putting 280TB of data onto 1,200 servers and making it all run smoothly to deliver response times in double digit milliseconds across the federated, sharded fabric cluster. Neo4j Architect Chris Gioran invites us to peek behind the curtain and tells the story of making this CEO challenge work. Interesting to see the fun we had with AWS and the approach we took to make it all happen.

And the best thing? You can try it for yourself. We put the code for generating data, provisioning machines, and the example app on GitHub. Just remember to watch your AWS bill!

How to Create Conditional and Dynamic Queries in Neo4j Bloom

bloom dynamic queries

Dan Flavin does it again. In one of his funny and deeply educational articles, he explores how to utilize the magic of and Bloom’s custom search phrases to create new search interactions you wouldn’t have thought possible. Truly mind-blowing if you’re a Bloom user. My favorite is running Cypher queries that return more data than Neo4j Browser can render.

Fetching Large Amounts of Data Using the Neo4j Reactive Driver: The Bloom Case

js reactive

Speaking of Bloom, in this blog post, Angeliki Komianou – engineer on the Bloom team – dives into the reasons and the how-to of using the reactive APIs of our JavaScript driver to optimize and speed up fetching large results from the database without the need to chunk manually.

NODES 2021 Videos: Graph Data Science

nodes data science

One popular track of the NODES conference two weeks ago was the Data Science track.

In the first part of the day we saw applications of knowledge graph talks covering COVID-19 from Saudi Arabia, patents, journals, and research at BASF to shell company activities. Practical demonstrations of integration with KNIME or using NLP for graph extraction from unstructured text to data transformation from relational databases were presented in the next section. We continued to apply graph data science on the meta-level to reconcile ontologies.

Then we saw a practical example of utilizing the Neo4j Connector for Apache Spark to allow data flow in both directions between tables and graphs. And as a highlight of the track, a full ML pipeline inside of Neo4j for Link Prediction used the GDS library. In closing, we enjoyed watching the exploration of a large dreams database with graphs, which is definitely one of the more unusual applications of graphs and data science.

You can find the video recordings of these exciting sessions in the youtube playlist.

Video: Ashley Faith Doing an Honest Review of Neo4j

ashley faith neo4j

In the fourth episode of her video series on Knowledge Graph Technology, Ashley reviews Neo4j. She talks about using Neo4j in a number of projects, the available tools, the awesome community, and what you can do with it. She’s joined by Dave Voutila who demonstrates a number of use-cases (like fraud-detection and personal data graphs) and how the graph model and tools (arrows, bloom, browser, neosemantics) support the mental model of business analysts/experts.

Ashley also presented this year at NODES. “Using Graphs to Support More Equitable Search” explores better ways for users to search and explore their data.

AuraDB Free Intro Videos

aura import csv video

Graph academy author Elaine Rosenberg guides you along your first steps with Neo4j AuraDB Free. In the first few videos, we create a database and load a CSV or a dump file.

She also prepared a Neo4j Browser guide that shows how to import CSV data that was exported from a relational database. Just run :play import-relational-data-guide in your Neo4j Browser and another one for data modeling: :play data-modeling-guide.

Integrations in the Cloud

graphstuff fm

Will Lyon interviewed David Allen in this week’s GraphStuff.FM podcast episode on all things data integration – Spark, Kafka, Tableau, GraphQL.

You should definitely subscribe to the podcast. So far, all episodes have been worthwhile listening to.

Fittingly, the videos from the “Connections: Graph Architecture and Integrations” are available to watch at your leisure.

  1. Keynote Session: Extending the Reach of Graphs by David Allen, Partner Solution Architect and William Lyon, Developer Relations Engineer, Neo4j

  2. Neo4j with Cloud-Native Data Mastering for the Enterprise by Anthony Deighton, CPO and Julia Neagu, Director of Analytics, Tamr

  3. Utilizing the GRANDstack: Neo4j and GraphQL in the Real World by Will Reynolds, Principal Software Developer, Hoare Lea

  4. Analytics & Graphs: Neo4j Connector for Apache Spark by Andrea Santurbano, CTO Larus BA Italy

  5. Best Practices and Performance Optimization with the Neo4j Apache Kafka Connector by David Fauth, Field Engineer, Neo4j

Links of the week

twitchverse fastrp

Tweet of the Week

My favorite tweet this week was by Will Lyon, who is sharing the content of his NODES talk “Using Cypher with GraphQL” as a Twitter thread.

Don’t forget to RT if you liked it too!