Americas

  • United States
abednarz
Executive Editor

Avaki: A start-up to watch

Feature
Apr 21, 20033 mins
Enterprise Applications

These young vendors offer fresh approaches for addressing today's enterprise network challenges, from setting up secure wireless LANs to virtualizing data center resources.

Company name: Derived from a French-Polynesian word that describes the equitable distribution of resources in a community.

Avaki

Company name: “Avaki” is derived from a French-Polynesian word that describes the equitable distribution of resources in a community.

Origin: Andrew Grimshaw founded Applied MetaComputing, Avaki’s predecessor, in February 2001 to commercialize a government-funded grid computing project he initiated at the University of Virginia. He relaunched the company as Avaki that June.

Funding: A $4.2 million first-round extension closed in February 2003, bringing total funding to $20 million.

Investors: General Catalyst Partners, Polaris Venture Partners and Sofinnova Partners.

Top executives: Tim Yeaton, president; Andrew Grimshaw, CTO.

Products: Avaki Data Grid, for providing access to geographically dispersed data sources; Avaki Compute Grid, for aggregating and managing processing resources; and Comprehensive Grid, a bundle of Avaki’s data-and-compute grid products.

Avaki got its start building software to aggregate and manage unused CPU horsepower, but the company is making its mark solving a different enterprise challenge: sharing distributed data among remote teams and with external partners.

Avaki, of Burlington, Mass., released its first product, Avaki Compute Grid, in the fall of 2001. This traditional grid software is designed to harness unused CPU cycles and dole them out to compute-intensive applications.

Customers took to a feature that let them serve data to users and applications over a wide area, inspiring Avaki to expand its focus to include distributed data access, says Tim Yeaton, Avaki president. The company initially integrated data-grid functions in Compute Grid, then released the stand-alone Data Grid software in December 2002.

“Data Grid came to be because customers told us that wide-area data access was a fundamental issue in their companies,” Yeaton says.

Avaki Data Grid gives users access to data that’s scattered across myriad corporate and partner systems, letting distributed teams collaborate without having to know where data is stored. Administrators can serve data from a single source to all users and applications without having to replicate or alter existing data stores. It’s an alternative to access technologies such as VPNs, FTP and demilitarized zones, Yeaton says.

No client software is required for Data Grid, which starts at $35,000. Data Grid sits on servers at each team location; firewall proxies let Avaki traffic pass through firewalls. The software handles usage rights and access privileges so administrators can limit what different users can access.

With Version 3.0 of Data Grid, introduced in December, Avaki ported the software to Java and Java 2 Platform Enterprise Edition to make it easier to deploy and integrate with existing IT infrastructures, Yeaton says. The company built in enterprise software capabilities, including failover support and the ability to integrate with existing directory services, such as Lightweight Directory Access Protocol (LDAP), Microsoft Active Directory and Sun One Directory Service, for user authentication.

These days, Avaki is trying to expand grid-computing implementations beyond traditional academic and research settings, and into the corporate realm. Targeted industries include life sciences, manufacturing, energy and petrochemical. The company’s 12 customers include Gene Logic, Infinity Pharmaceuticals and Pfizer.