Njini Granted $13M

Njini has picked up $13 million in funding to help it try and make its mark on the emerging data classification market.

Njini CEO David Jones says the U.K.-based startup will use the funding to enhance its Information Asset Management (IAM) software suite while expanding sales and support, and establishing more of a presence in the U.S. (See Njini's Out of the Bottle.) Jones says previous investors Accel Partners and Add Partners are back, along with a new lead investor he won't name until the round officially closes later this month. The B round brings Njini's total funding to $19.5 million.

"This gives us working capital we need to build sales and support infrastructure," Njones says, adding that the startup has fewer than 10 customers and 30 employees. He expects to grow headcount to 50 by the end of the year, in hopes of growing its customer base in a young market served mostly by startups.

IAM categorizes and tags unstructured data and lets businesses set policies to determine how many copies should be kept, where it should be stored, and who can access it. CTO and founder Phil Tee says the goal is for a significant product release late this year, with high availability features and CIFS compatibility among the additions.

With startups Abrevity, Arkivio, Index Engines, Kazeon, Scentric, and StoredIQ also shipping data classification products, analysts say the keys for the contenders are the way they differentiate their products and the partnerships they build with storage vendors. (See Scentric Gets Classified and Peekaboo, StoredIQ!.)Njini's differentiator is that its software runs on an inband appliance, which puts it in the data path. Njini says this allows it to classify data in real time instead of running in batches overnight.

Njini already has one large public fan in Virgin Mobile, one of the leading British mobile phone carriers. The company was among the first to test IAM and is getting ready to roll it out in production.

Keith Bennett, Virgin Mobile's senior infrastructure architect, says he expects Njini to save him about 20 percent to 24 percent storage utilization by storing duplicated files once. Bennett says the firm has about 500 users on shared drives, and many of the files stored are redundant.

"We're wasting a significant amount of storage," Bennett says. "We're also looking to move storage across tiers automatically. We were looking at implementing tiered storage for ILM, but couldn't find any storage products that did it. Njini places specific types of files on specific tiers."

Although data classification vendors bill themselves as ILM or compliance products, analyst Brad O'Neill of the Taneja Group says he expects even the successful ones will handle only part of the process within large storage environments."Njini, like all of these vendors, will probably end up more applicable to a subset rather than all of these usage cases," he says. "Their focus right now is in leveraging classification to optimize data placement across tiers based on policies."

O'Neill says the major vendors are likely to piece together different classification products, either by acquisition or partnership. For instance, Kazeon has an OEM deal with Network Appliance that gives it the lead in mind share if not actual market share. (See Kazeon Pairs With NetApp.)

"The classification space has attracted a lot of startup activity, and the partnering dances are in full swing with all the major vendors," he says. "I expect major vendors to use multiple tools to solve each task."

Jones says he is pursuing partnerships for those reasons, and one more important one: "Strategic partners are essential for any new technology. Here you're dealing with a corporation's key assets."

Barnett says that was an issue at Virgin Mobile, which usually sticks to proven technologies."We don't tend to implement the latest and greatest, and Njini is literally right out of the factory," he says. "We've done an extended proof of concept where we've tested the product to death, and then we'll roll it out to around 50 end users while still having the original drives mapped on original filers so we can revert back. There was a bit of trepidation from the board until we told them the way we're doing it."

Dave Raffo, Senior Editor, Byte and Switch

Organizations mentioned in this article: