​Canberra-based startup Instaclustr targets $3b big data market

With the copious amounts of data produced every day as its motivator, database-focused startup Instaclustr is positioning itself to take on the $3 billion big data market.

In 2013, Ben Bromhead and Adam Zegelin set out to tackle the data marketplace. Working from Bromhead's spare room in Canberra, the pair realised in their attempt to store data that they needed someone to manage it.

Bromhead and Zegelin already determined Apache Cassandra was the right technology to manage their marketplace, but as no one was offering a managed service for it, they started doing it themselves.

"We thought there may be other people facing a similar problem who may be interested in what we built and over the space of a week we put together a simple web interface, threw on some credit card billing, and pushed it out the door with a blog post and a message to a Cassandra mailing list," Bromhead told ZDNet.

Apache Cassandra is an open source distributed database management system, and is commonly categorised as a NoSQL database. Bromhead said it was created as a result of the number of limitations that were associated with traditional or relational databases like Oracle or MySQL.

Cassandra originated from Facebook, with the social media giant building the technology to run its messaging syste and. Mark Zuckerberg's team then open sourcing and released it to the community.

"When Facebook released Cassandra, a lot of companies saw the benefits of the scalability and availability it brought to the database world and we've actually seen some very big companies adopt this technology like Netflix and Spotify, but it's not just hi-tech companies but also traditional bricks and mortar stores like AT&T, Walmart, and ING Group," Bromhead said.

"Probably the crown jewel is actually Apple. [Apple is] a huge user of Cassandra with iTunes and the App Store's main data store being Apache Cassandra, so they trust this open source technology with these two product lines that generate a huge amount of profit or revenue for them."

Essentially, Instaclustr take the open source technology, package it up, and run it on a cloud environment. Bromhead calls this Cassandra-as-a-service.

According to the company, there has been a substantial shift in the industry, with companies employing open source software and NoSQL database technology, brought upon by the explosion in the amount of data that companies have had to deal with since the arrival of Web 2.0.

"With this increasing amount of data, companies more and more have to deal with large volumes of data that no longer fit on a single server or single computer," Bromhead said.

"Because there are so many companies that are dealing with big data and need to be able to solve these big data problems, this is a big market, this has been estimated to be about a $3 billion market by 2018.

"What Cassandra does is it takes this approach of taking very large amounts of data and splitting it up into very small components or very small portions of that data and placing it on individual servers so all these individual servers run as a single logical database, often called a cluster."

Instaclustr works with 100 customers globally, with startups, ad tech, education, social networks, and Internet of Things (IoT) the most common sectors. With Australian customers including Campaign Monitor, Maths Pathway, and Vero, Bromhead said the draw card for customers is that Instaclustr can typically provide the database management at a fraction of the cost it would take for a company to do it alone.

"What we first started offering was database-as-a-service but we've actually expanded into some broader technologies that really complement Cassandra," Bromhead said.

"We're primarily known as a Cassandra-as-a-service business, but in terms of the big picture we really are heading towards offering full big data-as-a-service."

Bromhead said that over the last few months the company has developed a managed solution for the Apache Spark analytics engine.

"This is just one of several other complementary technologies -- including Apache Kafka and Zeppelin -- which move us from our original offer of being a database-as-a-service provider to one that is big data-as-a-service.

Instacluster CEO Peter Nichol said staying in Canberra provides Instaclustr with a competitive advantage, thanks in part to the relationship the company has formed with the territory's universities.

"We've been able to grow a world class team of Cassandra experts, of technology experts who are as good as anybody in the world, but not as expensive as they would be in the middle of Silicon Valley," Nichol said.

When it comes to the topic of an initial public offering (IPO), Nichol said Instaclustr look at all options when it comes to funding, and that an IPO is just one the company is contemplating.