Intelligent Enterprise

I've been very impressed by ParAccel, a company that has lived up to their Intelligent Enterprise designation as a Company to Watch. They sell a columnar, MPP DBMS that has won strong market visibility in the two years since the company's founding. The industry experience of ParAccel executives, gained at companies including Oracle, Teradata, Netezza, and DATAllegro, surely plays a part in the company's success, and so must two technology antecedents: an earlier columnar DBMS, Clareos, and the PostgreSQL open-source RDBMS.

On the PostgreSQL front, all I can say is that I share others' appreciation for the number of commercial products the project has spawned, benefiting from its use-it-however-you-want BSD software licensing. Vendors with PostgreSQL-rooted products also include Dataupia, EnterpriseDB, Greenplum, Netezza, and Truviso among others. (A thank you to Simon Riggs of 2ndQuadrant for guidance on this point. And note that of those companies, only EnterpriseDB distributes an open-source product, Postgres Plus, which adds MPP and other capabilities to base PostgreSQL. A Truviso press release does state, "Truviso-sponsored improvements are expected to be included in the next release of PostgreSQL, scheduled for early 2009.")

Clareos's CrossCut product wasn't the first column-store system, but I believe it was the first columnar MPP system, designed for parallel operations. The ParAccel link is Director of Engineering David Steinhoff, who founded Clareos in 2000. I spoke to Steinhoff at length earlier this year. He was forthcoming about lessons learned from his first column-store start-up, which folded in 2006: "At Clareos, we had a technical vision and a market vision that didn't fit seamlessly into the market space we targeted." That impatient investors launched a lawsuit against the company "was a painful distraction for senior management." While Steinhoff said the lawsuit "didn't hurt us operationally," it must have diverted resources that otherwise could have gone into the marketing and business development activities that could have successfully defined a column-store DBMS category. That visibility was created by the launch of ParAccel-rival Vertica in 2005. Michael Stonebraker's star power, harnessed to promote Vertica, defined a market category where Clareos (and other established columnar vendors) did/could not.

Steinhoff continued, "At Clareos, we focused on technology. We wanted to have the absolutely best technology and engineering. ParAccel is more focused on delivering a complete solution." ParAccel has found opportunity, according to Steinhoff, in "shops [that] have SQL Server, Oracle, possibly DB2. Others we see are Vertica, DATAllegro, Greenplum: the new generation, the parallel generation."

ParAccel VP of Marketing Kim Stanick reports recently closing new deals with data and analytic providers in the face of competition from other MPP and columnar vendors. Stanick says that key customer decision points include flexibility, feature-set breadth and richness, and price/performance. According to her, one prospect is replacing an "underperforming" conventional-RDMBS-reliant solution because "long running... queries delay delivery of analytic products to their customers." According to Stanick, "since launch, ParAccel remains unbeaten in POC (proof-of-concept) trials."

It should be noted that ParAccel has not been rigid in their market positioning. The company first led with a query accelerator for SQL Server and Oracle, but they modified their messaging to focus on the value their products can deliver as free-standing database engines. Philip Howard of Bloor Research published a helpful overview on this and related points last October. And they have entered into noteworthy alliances with vendors that include Informatica, Talend, EMC, and JasperSoft.

ParAccel's has benefited, then, from strong technical heritage, industry experience, careful but adaptive positioning, and a serious head start afforded by a robust, open-source code base with business-friendly license terms. We in the BI/DW worlds can look forward to more good things from them and the whole of the MPP/columnar DBMS solution space.