Big data, little companies: These six startups want to disrupt the data world

NEW YORK — The Strata + Hadoop World conference attracts the biggest names in big data — think Microsoft, IBM, SAP, Facebook, and Cloudera — but it also gives some smaller, emerging players a chance to shine.

On Monday night, VentureBeat had a chance to chat with some of the Strata startup showcase finalists to learn about how they’re approaching big data differently. (If you have a great offering and your company isn’t featured here, it may be because we didn’t have time to talk to everyone. Readers, see the full list of showcase finalists here.)

With that said, here are six promising startups we met at Strata.

Editor’s note: Our upcoming DataBeat conference, Dec. 4-Dec. 5 in Redwood City, will focus on the most compelling opportunities for businesses in the area of big data analytics and beyond. Register today!

Sqrrl

Sqrrl offers a NoSQL database for private enterprises that relies on technology created by the National Security Agency.

The company wants to convey one thing above all else: its database is secure.

“We are the most secure, scalable, and flexible NoSQL database out there,” Ely Kahn, Sqrrl’s vice president of business development, told VentureBeat. “And I’d probably put those in that order in relation of importance to our customers.”

In 2008, NSA director Keith Alexander decided to build a powerful, secure database that could cull data from disparate silos. Called Apache Accumulo, the database was always meant to be open source, but it took the programmers behind the project three years to convince the NSA to open it up to the larger ecosystem of Hadoop developers.

Sqrrl tags every key value pair — the most granular level of storage — with a unique security label that dictates who can access that piece of data at the application layer. But it doesn’t stop at key values: Sqrrl also features graph store and document store capabilities.

The company already has six paying customers: two government agencies, two major banks, a health care company, and a telecommunications company, according to Kahn, who declined to provide the organizations’ names.

Founded in 2012, Sqrrl has offices in Cambridge, Mass. and Baltimore, Md. It raised a $5.2 million funding round from Atlas Venture and Matrix Partners last week. Sqrrl won the audience choice award at the Strata startup showcase.

Metric Insights

Is this the one ring — er, business intelligence tool — to rule them all?

Metric Insights‘ software ties into data sources like Hadoop, MongoDB, and Cassandra (among many others) and displays the resulting metrics in a customizable, predominantly visual web interface.

The problem with business intelligence today, said Metric Insights COO Steve Mock, is data overload.

“There’s way too much data and way too many dashboards,” he told VentureBeat. “The bigger and messier a network is, the more appealing our system is. Our customers typically have two or more business intelligence tools.”

Some of those customers include Barnes & Noble, United Online, and HomeAway, which pay Metric Insights an annual subscription fee per user.

Metric Insights is cash-flow positive, according to Mock, and has only raised money from private investors so far — though it’s in the middle of raising another round now.

Metric Insights snagged first place in the Strata startup showcase. Congratulations!

Alpine Data Labs

Alpine Data Labs wants to make data analysis accessible to everyone in the enterprise, not just the data scientists.

The San Mateo, Calif.-based startup this week launched Alpine 3.0, the first full-fledged version of its analytics solution for Hadoop. The browser-based software is primarily visual, replacing scripting with drag-and-drop actions.

“When people ask what our language is, I tell them, ‘the language of the mouse,'” said Bruno Aziza, chief marketing officer at Alpine Data. “We want users to rule the math, not rule the code.”

There are obviously tradeoffs to making data analysis primarily visual, but Aziza thinks it’s okay if Alpine loses the academic researcher or advanced statistician by appealing to less technical folks in the enterprise.

“Let’s go after the 80 percent first, and after the 20 percent later,” he said.

Havas Media is using Alpine to figure out which social platforms are the most effective for ad campaigns. Alpine enables the sales and marketing people to interact with the data directly, rather than relying on data scientists.

“And that’s the game changer,” said Aziza.

Instead of moving data to a separate analytic server, Alpine’s software takes advantage of Hadoop clusters’ massively parallel processing power. In other words, data gets processed quickly because it doesn’t have to go anywhere.

Alpine Data Labs has been working on its accessible analytics platform for the last 18 months. It currently has 45 employees and counts Barclays, Equifax, EMC, and Disney among its customers. Alpine raised a $7.5 million funding round in May 2011.

Above: The Strata startup showcase after 8 PM.

Image Credit: Eric Blattberg / VentureBeat

JethroData

JethroData is an index-based SQL engine for Hadoop that promises faster queries than Impala, HAWQ, or HIVE.

“As data comes in, we index it on the fly and store it in [the Hadoop Distributed File System],” explained JethroData CEO Eli Singer. “And now we know the results when you need those records.”

So rather than scanning the entire database for a query, it’ll pull specific records related to that query — which it already knows, because they’ve been indexed in advance.

In other words, JethroData is really good at picking the needles out of digital haystacks. If you search for the whole haystack, it obviously won’t run as quickly.

“If you look at the history of databases, this is not a new concept, but it’s new to Hadoop,” said Singer. “The more you drill down in your query, the more effective Jethro becomes.”

The Israel-based startup closed a $4.5 million round led by Pitango Venture Capital in February 2013. The product is currently in closed beta; Singer forecasts a release in the beginning of next year. The company currently has 12 employees.

Hazelcast

Hazelcast, an in-memory data grid, began its life as an open-source project in 2008 and was later spun off into a company.

It’s designed to tackle one of the biggest problems for Java developers: preventing applications from drowning in data.

Above: Hazelcast’s Miko Matsumara

“Distributing caching and processing are very easy for us,” said Miko Matsumara, Hazelcast’s vice president of marketing. “Your distributed cache is embedded in the application, so it can scale elastically with the demand — and it keeps applications separate, so one traffic spike doesn’t present problems for the rest of the applications.”

In an era where everything comes as a service, Matsumara joked that Hazelcast offers caching-as-a-service.

Hazelcast remains open-source. Matsumara estimates it has “a few dozen” outside contributors. It’s already racked up some big supporters, including Apple, Firefox, AT&T, and Cisco.

Hazelcast raised $2.5 million in a funding round led by Bain Capital this September. The founders recently relocated to Palo Alto, CA, but the majority of the developers remained in Turkey.

Mevoked

Mevoked was the last startup we had the chance to talk to before a man with a bell shepherded everyone out of the area. It’s also the only consumer-facing platform on this list.

The San Francisco, Calif.-based company tracks children’s online behavior to provide parents insight into their mental state.

“We can look at early indicators of depression, obsession with violence, and risk of suicide based on online activity,” Mevoked founder and CEO Arun Ravi told VentureBeat. “We’re trying to intervene before that happens and empower parents to act.”

Above: Mevoked provides parents with daily updates on their child’s well-being.

Most children won’t take kindly to being monitored, of course, but in most cases the parents are the ones paying for their devices and web services, so Ravi thinks they have a right to know what their kids are up to online. But to mollify privacy concerns, Mevoked is working on a version of the platform that doesn’t show specific content.

Mevoked released its Chrome extension last week, which costs parents $5 a month to enable. Add-ons for Twitter and Facebook are coming shortly.

Ravi knows the service may be controversial — and he definitely knows that kids will hate it — but his motivations are pure: he wants to prevent mental illness in kids and teens.

“I had a friend who committed suicide when I was young, and I remember his parents saying, ‘I wish we knew, I wish we knew’,” said Ravi. “If we can reduce teen suicide by five percent, it’s worth it — because right now, everything people are throwing at it is not really working.”

Mevoked currently has four employees and is raising a seed round.