On June 17th, 2021, Neo4j, a graph database company, announced a $325 million investment led by a $100 million investment by Eurazeo and joined by new investors GV (previously named Google Ventures), DTCP, and Lightrock as well as existing investors Creandum, Greenbridge Partners, and One Peak. Eurazeo is private equity company with over 15 billion Euro in Assets Under Management as part of a larger investment portfolio of over 22 billion Euro. With this round, Eurazeo Managing Director Nathalie Kornhoff-Brüls joins the Neo4j Board of Directors.
This monster funding round speaks to the confidence that investors have in the future of Neo4j. But in this particular instance, Amalgam Insights believes that this large funding amount is especially important because of what it means for breaking the status quo of enterprise analytics.
Analytics and data management in the business world have been built around the relational database focused on controlling and governing individual data inputs. This fundamental framework has been very useful in creating an environment that can be configured to present a single shared source of truth. However, it is not especially good at supporting and processing data relationships, which is a challenge in today’s data environment as data grows quickly and data relationships increasingly represent some level of transaction or behavior aligned with a business activity that needs to be tracked or analyzed in near-real-time.
In addition, the hype regarding artificial intelligence and machine learning has finally crossed over into practical reality as the toolkits for operationalizing models have reached mainstream availability. Even as enterprises may not fully understand machine learning, but they can easily purchase access or use open source projects to access the data management, model creation, storage, and compute capabilities needed to support machine learning projects. But for companies to fully execute on the promise of machine learning, they need to create more efficient relationship-based data environments that allow models to be tested and to provide results. Building relationship-centered data is part of what I originally called the Battle for Context when Amalgam Insights was first founded.
And now four years later, Neo4j has a chance to deliver on this challenge for context at a global scale. Neo4j has been a graph data leader for years, especially since it started back in 2007 before the need for graph database management was fully clear to the enterprise market at large. Since then, Neo4j has been a stalwart in its market education of graph data. But it has fundamentally been fighting a status quo where companies have been either unwilling or unable to translate their key transactional data environments into the relationship-based models that will be necessary for broad machine learning. With this round of funding, Neo4j finally has a chance to conduct the volume of marketing and sales needed to educate the data and analytics audience. In contrast to other large rounds of funding announced in the data world, such as Snowflake’s $479 million round in February 2020 or Databricks’ $1 billion round in February 2021, Amalgam Insights believes that Neo4j’s funding round serves a slightly different purpose.
Those previously-mentioned funding rounds were all seen as final rounds of funding before an upcoming IPO with participation by software vendor partners in their ecosystem. In contrast, Neo4j both has a more foundational opportunity and challenge in that graph should be the foundation of enterprise machine learning and relationship-driven data environments, but the ecosystem and platform maturity are still not quite where the data warehouse market is. Amalgam Insights sees this round as being more similar to DataRobot’s $270 million round raised in November 2020 which allowed DataRobot to continue acquiring companies and building out its platform to fit enterprise challenges.
Ultimately, the goals that enterprises should associate with graph data are the combinations of unlocking relationships within data that will take orders of magnitude in time, money, and skillsets to discover in relational data as well as the opportunity to unlock tens to hundreds of millions of dollars in value through ongoing machine learning and artificial intelligence operationalization opportunities that have already been identified but cannot run at high-performance levels without a better data environment. The acquisition and use of graph databases is a technological bottleneck that will prevent enterprises from fully unlocking AI and we are only now reaching a point where the understanding of relationship data, training data, machine learning feedback, and transactional data is sufficient for business managers to understand the value of dedicated graph databases rather than simply placing a graph structure on relational or multimodel data.
Recommendation to the Amalgam Insights database and analytics community
At the very least, start learning about graph data structure as combinations of edges, vertices, and relationships as well as linear algebra to gain an understanding of how graph data differs from the standard high school algebraic logic of relational databases. Yes, learning math and a new set of data relationships is not as easy as downloading a library or learning a new software functionality. But graph relationships are a fundamental change in the way that data will be managed over the next couple of decades and there will be a great deal of work needed both to ETL/ELT relational data into graph databases as well as to manage graph databases for the upcoming world of AI replacing aspects of standard business analytics.
If your organization is looking at relationship analytics or machine learning initiatives beyond a single project, look at Neo4j, which currently has a dominant position as a standalone graph database and is available as open source under GNU General Public License (GPL v3).
And if you have questions about the current state of Neo4j or are trying to bridge gaps from BI to AI in your organization, please contact email@example.com to schedule time to speak with our analysts. We look forward to serving you in our continued role in helping you to understand the future of your data.