As more information becomes digitized, and more relationships are created between objects (data), the much larger the data set for computers to analyze. This means getting an answer to a question can take a lot more time than our impatient selves can handle.
And, this means structuring the way data is stored is more important than ever. To this end, these days, the move toward graph databases is starting to make a whole lot of sense for more and more companies. It's no wonder, therefore that Neo Technology, creators of Neo4j graph database, announced Friday it raised $11 million in Series B funding.
Sunstone Capital led the round, with participation from all of Neo Technology’s current investors, including Fidelity Growth Partners Europe and Conor Venture Partners. Neo Technology has now raised a total of $24 million.
Neo Technology plans to put some of the funds toward new features that will make it easier for average developers to consume the product. This is something the company has not done a good enough job of to this point, Emil Eifrém, CEO of Neo Technology, said in an interview.
There will be new usability features, including a new query language, something that Eifrém says graph databases have not had in decades.
"We're looking at a number of things that we'll be rolling out over the next year, including a curated learning experience, improvements to our Cypher query language--which is proving very popular because of its intuitiveness, a number of improvements to our remoting experience--including upgrades to the current REST API... and a few surprises," Eifrém said.
What is a graph database?
For the last 20 to 40 years, Eifrém said, databases were boring, since there was no choice about what database to use; they were all relational databases. That has begun to change in the last few years, though, with the rise of alternative databases, like the ones that use graphs.
The difference between relational and graph databases is that relational graphs have to search all of the data in order to find anything that meets the search criteria. For larger sets of data, this can take a very long time.
A graph database, on the other hand, will search records that are directly connected to other records. It can even be given a limit on how many "hops" it is allowed to make, so that it will ignore everything more than that number of hops away. Graph databases model computer networks, social networks, or anything that links data together, and then calculates the shortest routes between the two points. This becomes especially important when dealing with large amounts of data.
To put in simple terms: a relational database is like going through every page of a book until you find what you were looking for, while a graph database is like looking in the index and then going directly to what you need.
So, unlike the old databases, graph databases are optimized for connecting data, and that is what apps being developed now need. Eifrém says that graph databases are 1000 times faster than their older counterparts.
Eifrém gives the example of Facebook, which he said was using a tabular database, but was experiencing all kinds of performance problems, until it began using a graph database, which allowed people to connect with each other. By using a graph database, Facebook was able to better trace the proximity of stated relationships between two people.
A little about Neo Technology
San Mateo-based Neo Technology was originally founded in Sweden in 2000, and only set up U.S. operations in 2011. At the time it had only one global client: Cisco.
Now, only a year later, the company has 20 Global 2000 customers, including Cisco, Vivendi, Deutsche Telekom, Lockheed Martin, BAE Systems, Telenor, LexisNexis, Intercontinental Exchange, Adobe, and Pitney Bowes. Eifrém says the company also has 50 smaller clients, including Squidoo, FiftyThree,GlassDoor, Teachscape, Gamesys, Viadeo, MoviePilot, InfoJobs, and Woozworld. It also has thousands of open source users who use the free version of the software.
There is also a paid version of the software, which comes with additional features., including offer support, and a commercial license, as opposed to open source.
"The Enterprise Edition, designed for high-availability production use, and our mainstay offering for paid users, includes a number of high availability features, such as online backups and high-availability clustering," Eifrém said.
Every quarter this year, Neo Techology's bookings have more than tripled as compared to the sameperiod last year. Total bookings in 2012 are up 440 percent over 2011.
What sets Neo Technology apart, says Eifrém, is just that they are bigger than all other graph database companies combined. He says that this is simply because they got started early, even before Facebook announced its social graph.
The philosophy behind Neo Technologies is twofold.
One, it is not just enough to state what the data is, but to make sense of it, Eifrém says. Other companies are good at obvious things, but not at drawing conclusions from their data.
The second part of the philosophy is that it is the tech people that should distribute the software. Often, he said, managers and directors will buy software that their tech people hate, but which they have to use because that is what they have.
Eifrém gives the software to tech people for free, do they adopt it because they love it, and later they can ask managers to purchase it for the company. He doesn’t want to sell the product top down.
Of course, this does not preclude the company from then pitching the product directly to the managers once their tech people have recommended it. Eifrém just says he wants the support of the “ground troops” first.