Zillabyte Relaunches To Provide Infrastructure For Data Apps

Zillabyte, a Y Combinator-backed startup that previously billed itself as a “Pandora for sales leads”, is officially relaunching today with a new focus on becoming “a cloud platform for data analysis.”

The goal, according to co-founder and President Roger Huffstetler, is “empowering the lay developer to use distributed computing.” In other words, if a developer wants to build an application that involves data analysis, rather than worrying about the infrastructure, they can use Zillabyte to “abstract all those levels of infrastructure and focus strictly on the model you’re trying to build.”

Huffstetler compared Zillabyte to Amazon Kinesis and Google Cloud Dataflow — but he described those tools as “my distributed computing tools for my distributed computing engineer,” whereas Zillabyte is aiming for a wider audience of develoeprs.

That’s not quite as big a shift as you might think from the company’s previous efforts, which involved recommending sales leads who were similar to a business’ existing customer base. Huffstetler said that to provide that kind of analysis, Zillabyte already had to build this infrastructure layer for itself, and ultimately, the team decided to focus on the platform rather than the sales application.

Asked for more detail about why Zillabyte made the switch, Huffstetler added:

There are a few things. First, we started to think a lot about our key passions and experiences, and, to borrow a phrase from Chris Dixon, we really knew our founder-market fit was in scaleable, easy-to-use infrastructure, given our backgrounds at Google and Twilio. [Huffstetler was previously a sales manager at Twilio, and CEO Jake Quist was a software engineer at Google.]

Second, we started to look at some competitors in the lead scoring and lead generation space, where we started, like Infer, Lattice Engines, and Mintigo. These are all great, venture-backed businesses, but then we asked ourselves: What if we could help power those companies?

Zillabyte allows users to work with both open datasets and their own private data, and also offers “components” that are basically packaged algorithms to cover tasks like geocoding, fraud analysis, and web scraping. The platform currently supports Ruby, Python, and JavaScript — to give a sense of how easy it should be to use, the very first thing you see at the top of the company website is code for creating a word-counting app in all three languages.

As examples of what people can do with the platform, Zillabyte’s blog has featured apps that predict changes in the price of metals, track your competitors, and, yes, find sales leads.