Rocky Mountain High OLAP
When a group of Colorado community colleges wanted an affordable OLAP solution, it turned to Microsoft. And Microsoft turned to integrator Len Silverston. "Why can't reports summarize data the way we want it?" Solutions integrator Joe Misiaszek has tried to answer that question for years. His programmers pounded out reports as fast as they could, but it was never fast enough.	"Understand the customer's needs and get the users to buy into the design. That's key to success for any data warehouse." Len Silverston, Quest Data Solutions, LLC

Misiaszek is manager of application software and IT at Community College Computer Services (CCCS), an organization that maintains computers for 13 Colorado community colleges, as well as the state board that governs them. These systems contain enrollment and financial records for 2,500 students and 3,500 staff members. The reports that emerge from the files were state-of-the-art 10 years ago, when the system received its last major upgrade.

Last year Misiaszek's boss, Don Williamson, vice president of IT at CCCS, encouraged him to investigate system upgrades. The more I looked at available decision-support solutions, the bleaker the picture became, says Misiaszek. I couldn't cost-justify anything I saw. Then he heard about the new OLAP engine that Microsoft Corp. was bundling with SQL Server 7.0.

Microsoft offered the features Misiaszek needed at an affordable price, but he had never implemented a data warehouse. Microsoft referred him to Len Silverston, president of Quest Data Solutions LLC in Greenwood Village, CO, a $9-million solutions-integration company specializing in databases and data warehousing.

In February, CCCS hired Quest to jump-start and coordinate the project. Silverston's first task: Understand the customer's needs and get the users to buy into the design, he says. That's key to success for any data warehouse.

Subsequently a committee of powerful decision makers was formed, representing every major constituency in the CCCS organization. Under Silverston's guidance, the committee crafted a design that everyone could live with. (As with most data-warehousing projects, Quest's design added new reporting capabilities that augmented, but did not replace, the existing database.)

But the pilot test clinched the deal. OLAP presentations are very visual, and if a picture is worth a thousand words, it also can be worth many thousands of dollars to Silverston.

We brought the committee members into our lab and gave them the URL to the data warehouse, says Silverston. I stood back and watched unaided users generate year-to-date reports every bit as good as the ones programmers create by hand. 'When can I get this on my desk?' was the only question they asked.

Misiaszek and his three teammates were largely responsible for the production phase. Len kept us focused on the plan and timetable, says Misiaszek, but we did most of the actual development.

A prototype was completed in April 1998, and by late August the production system was up and running in four schools and the state board. Misiaszek expected to have all 13 schools online by the end of September.

What he's truly excited about, however is implementing this solution for less than one- of the price that most other vendors quoted. The sum total: 14 Digital Equipment Corp. AlphaServer 2100 quadprocessor servers, 14 copies of SQL Server 7.0, and Microsoft's OLAP server-all for less than $225,000.

What's Misiaszek going to do with the time he gains from having users crank out their own reports? The state board is already asking how soon I can add a data warehouse to the student information system.