Big data, big growth

Open source NoSQL companies are making headlines for investment figures, but they’re offering knowledge and building communities too

We’ve looked at small and exciting FOSS start- ups in the last two months, but not all open source start-ups start small. Some of them grow quite big, pretty quickly. If there’s one area generating buzz, where free software tools and the companies behind them feature very strongly, it’s big data, and behind big data (unless you have money to burn) is some flavour of NoSQL.

NoSQL databases aren’t just for companies heading for a billion customers. NoSQL tools enable companies to capture and analyse more than ever before, at a fraction of the cost of traditional solutions. This long tail of diverse new users has pushed the growth of companies behind NoSQL solutions as much as the behemoth customers who grab all of the headlines.

In the last couple of years the companies behind Riak and Redis have both raised around $10m (around £6.2 million). Data Stax, who build their enterprise product on top of Apache Cassandra, have raised ten times as much – while MongoDB lives up to its name with humongous growth – and substantial VC funding has also gone to companies associated with Hadoop, Couchbase, and MongoDB. But the buzz around (open source) NoSQL isn’t just about investment, it’s about knowledge too.

Cassandra has a real community of users pushing interesting solutions and building high value start-ups. From Boxever in Dublin – whose airline and travel big data and personalisation platform has brought them enough growth to double their employee numbers – to SoundCloud in Berlin, tech start-up companies who use FOSS NoSQL for pioneering analytics are hosting Cassandra MeetUp groups and Big Data Beers. This year SoundCloud, for example, hosted Data Stax CTO Jonathan Ellis, who talked about a new Cassandra release, and at Boxever’s HQ, you could hear Frederik Skogberg from Paddy Power talking about a scalable journaling system built on Cassandra.

Amidst all the exciting technology and occasionally silly sums of money, it’s important not to lose sight of this being a story of free and open source software not just succeeding, but doing so against established proprietary companies with very deep pockets like Oracle. They are doing so with, in many cases, entirely free and open software and not just some value-added “enterprise” products built upon open source software.

One company that stands out from this crowd is Hazelcast, who this September raised an extra $11m (around £6.8 million) in funding. What makes them different is that their product (yes, an open source NoSQL one) is an In Memory Data Grid (IMDG) and is one of their own development. This enables very large-scale, transactional In-Memory Computing applications, with great potential for emerging real time data applications.

Two years ago, the young Turkish start-up already had Cisco, Ericsson, and Mozilla as customers for its scalable caching solution. Last year, as well as moving the company to Silicon Valley, they took a Lego-cased, 48-way Raspberry Pi cluster to Oracle’s JavaOne conference to demonstrate their elastically scalable KeyValue store, which has been benchmarked as rather faster than Cassandra.

Recent funding is aimed at expansion to bring them head-to-head against extremely expensive solutions like Oracle Coherence, VMWare Pivotal Gemfire and Software AG Terracotta, which opens IMDGs to many more industries for the current well-heeled customers like investment banks, in the same way that other NoSQL providers are building whole new industries.
There’s certainly a lot of money going into open source start-ups at the moment – but is this always a good thing? Next time we’ll be looking at start-ups which reject the equity funding route and why they choose to do so.