Start-up manages unstructured data

Posted on March 15, 2005

RssImageAltText

By Heidi Biggar

This week, the industry got its first glimpse of Kazeon, a self-described provider of unstructured information management products for the enterprise.

Kazeon is one of a growing number of companies, which include specialized start-ups, e-mail archiving vendors, storage vendors, and content management players, that are trying to tackle the growing management problem of unstructured data types (e.g., e-mail, files, etc.).

"We're going to see a lot of activity in this space throughout 2005," says Brad O'Neill, senior analyst with the Taneja Group consulting firm. "Every major vendor is waking up to the reality that 'content-aware classification' accelerates a range of business processes, such as compliance, corporate security, and consolidation."

Kazeon, a two-year-old start-up in Mountain View, CA, is expected to share initial details about its upcoming product launch, as well as information about its expanded beta program, at the Storage Networking World show next month. The company is backed by $17 million in funding.

"Kazeon is a player in a new category of advanced classification and control software solutions," says O'Neill. "It uses content-aware technology to classify and control information, not just external metadata."

To be able to truly leverage the value of unstructured data for various strategic business purposes (e.g., compliance, litigation support, security, and consolidation), users need to know a lot more about the content of the unstructured data than metadata alone can provide.

"They need to know what is in the files, how the files are accessed and by whom, and how they are secured—and they need to be able to take control of the content and take automated actions based on the content of the files," says Troy Toman, Kazeon's vice president of marketing.

Taneja Group's O'Neill says that while a variety of content management, e-mail archiving, data classification, and data movement/policy management applications exist today that help users manage certain aspects of their unstructured data archives, true data classification and policy management are still several years away.

Nonetheless, O'Neill says vendors such as Arkivio, Deepfile, and Kazeon are taking significant steps toward this end, developing information management technologies that "give administrators much more granular visibility and control over what gets archived on an [object-based archiving platform]."

Kazeon says its focus is to infuse "content-awareness" into object-based archives, as well as other systems and applications, to help users manage growing unstructured data stockpiles.

While products such as EMC's Centera Seek go a long way in validating the market in which Kazeon will play, Kazeon's Toman claims that those products are generally "too passive." As for content management applications, Toman contends that they can require too much overhead and can be complex to manage.

Kazeon has designed its first product—an appliance—to be complementary to, not competitive with, products from EMC and others.

For example, the Kazeon appliance can be used to "comb" a user environment to find files that should be stored on EMC's Centera platform, or other vendors' object-based archiving systems. "It provides visibility into and control over what unstructured content can get archived where and when," explains O'Neill.

Kazeon describes its platform as an "advanced data classification system." The system, or appliance, creates and stores abstracts for each file in the network; the abstracts are stored across the appliance's "distributed repository," which can reportedly scale to billions of objects. A Kazeon API facilitates communication with other applications and systems, allowing them to directly contribute data to, and make requests from, the Kazeon platform.

The Kazeon platform also supports a number of integrated services, including storage search, reporting, auditing, archival, and policy management.


Comment and Contribute
(Maximum characters: 1200). You have
characters left.