ZoomData raises $17M to become the visualization layer for big data

0 Comments

ZoomData is fairly unique in the world of big data startups, headquartered more than 2,500 miles from Silicon Valley in Reston, Virginia. Yet the company has been attracting attention from some big Silicon Valley names, including venture capital firm Accel Partners, which led the $17 million series B investment round that ZoomData announced Monday.

Some folks might remember when [company]ZoomData[/company] launched nearly two years ago, promising to remake data analytics on the iPad with connections to big data stores such as Hadoop and gesture-based visualizations designed to be manipulated. It was focused on streaming data, and the application let users rewind and fast-forward through views of their data as it hit the application, so they could see real-time data, look back at five minutes ago, then walk back up to the current view all with a few swipes.

It was a promising approach, but a limiting one, ZoomData Co-founder and CEO Justin Langseth said. The company learned pretty quickly — when customers were sending it terabyte-sized hard drives to upload, and it was sending customer boxes of iPads for their data analysts — that historical data and desktops still matter very much, especially when you’re talking about big data.

A screenshot of ZoomData for the iPad.

A screenshot of ZoomData for the iPad.

But it was also a fortuitous approach, if only because of the technology that ZoomData built in order to do real-time visualization across large datasets. According to Langseth, the company stumbled across a way to turn its real-time technology onto historical data, as well, as it was uploading one of those hard drives a customer sent over. They uploaded it to Hadoop, put Cloudera’s then-new Impala SQL engine on top of it, and realized they could create timeline views of historical data “effectively using that exact same approach.”

The approach, essentially, involves stream processing the results of a query and displaying the results in a “sketch view” that gets sharper as more data is processed, not unlike the experience of watching the first few seconds of a streaming move. Users start to see their data almost immediately and get a good sense of what’s going on, and a short time later they see it all.

Langseth said he showed the technique to [company]Cloudera[/company] Chief Strategy Officer (and then-CEO) Mike Olson, who was impressed enough to form a relatively strong partnership with ZoomData. ZoomData has 20 paying customers, Langseth said, and a good number of them are also Cloudera users connecting ZoomData to their Impala or HDFS stores.

The ZoomData architecture. Source: ZoomData

The ZoomData architecture. Source: ZoomData

However, ZoomData isn’t tied to Cloudera or even Hadoop when it comes to the big data backends it supports. The company’s product also supports a collection of popular stores that includes MongoDB, Spark, Amazon Redshift, ElasticSearch, Solr and various relational databases and streaming APIs. Supporting a wide variety of data stores is critical if ZoomData is going to succeed in its lofty goal of unseating [company]Tableau[/company] as the go-to choice for analysis and visualization on top of these environments.

Langseth thinks the ability to do real-time and historical analysis in the same interface will be a big help in making this happen, but said the real key might be a focus on user experience. One big part of that is making sure the experience is equally as rich whether someone is manipulating a chart with their fingertips or a mouse.

“We’re really trying to make [analzying big data] even easier and even more beautiful,” he said.

NEA, Columbus Nova Technology Partners, Razor’s Edge Ventures and B7 also participated in ZoomData’s latest investment round. The company announced its $4.1 million series A round in June 2013.

Comments are closed.