• Home
• About Us
• Oracle BI
• Software
• What We Do
• Careers
• Links
• Contact Us
• Sitemap

Contact Details

Rockport Software Ltd,
126 Fairlie Road,
Slough,
SL1 4PY,
Berkshire,
U.K.
Tel: +44 (0)1753 577201
Fax: +44 (0)1753 577202
Email:

Click on the links below for additional information about Rockport Software

• About Rockport Software
• »  Rockport Management Team
• »  Rockport Company News
• »  Oracle Magazine Article

Oracle Magazine - July 2002

COVER STORY Enlighten By Alan Joch

Tight integration of core BI technologies within the Oracle database kernel and Oracle9iAS means you understand your business as never before.

When legendary gangster Willie Sutton was asked why he robbed banks, he replied, "Because that's where the money is." Business intelligence users may find themselves adopting some of Sutton's commonsense logic. Now, if asked why they're running their online analytical processing (OLAP) queries and data-mining algorithms directly against Oracle9i Database, these BI users might answer just as matter-of-factly, "Because that's where the data is."

Relational database management systems (RDBMSs) have long been the secure data bank for critical corporate information about customers, markets, financial performance, and other key business indicators. But marrying this data with the right slicing-and-dicing tools—whether to better understand current business conditions or to forecast future business—has traditionally required time-consuming data movements and tricky technical maneuvers. More often than not, data warehousing specialists had to take large chunks of data out of the finely tuned RDBMS and move them into a separate OLAP or data-mining engine. Only then was the data ready for analysis.

Fortunately for business managers who need access to real-time data—and for the IT staff that support them—BI systems have gotten out of the data-moving business. With the recent introduction of Oracle9i Release 2, Oracle now offers a suite of fully featured, tightly integrated BI capabilities that all reside within the Oracle9i relational database kernel. The result: All of the server capabilities needed for BI exist within the single Oracle Database engine, so that advanced OLAP and data-mining algorithms are built in to the database server.

Blending relational and multidimensional data into the same database engine could end the discussion on what's best for BI: relational OLAP (ROLAP), multidimensional OLAP (MOLAP), or hybrid versions (HOLAP).

"All of that debate goes away. You can leave the data where it is, have it where you want it, and build really powerful BI applications on top of it without having to worry about where the data is located," says Andy Johnson, consulting director with Rockport Software, a business intelligence and data warehousing consultant and software developer based in Berkshire, U.K.

And in the process, companies can devote more resources directly to their core business, rather than worrying about the underlying BI technology.

Integration Advances

This level of integration between RDBMSs, OLAP engines, and data-mining servers hasn't always been possible. The cornerstone for the current technology was laid when Oracle8i introduced object-oriented capabilities to Oracle databases, and with the introductions of Oracle9i Database and the BI tools built into Oracle9iAS, BI integration comes to full bloom.

Early reports from longtime Oracle technology users show they appreciate the potential of BI integration. "Oracle's data warehouse capabilities are huge advances," says Matt Rhoades, team leader, advanced technologies development, of Henkel Consumer Adhesives in Avon, Ohio. Rhoades manages a 1.2-terabyte data warehouse built on Oracle8i Enterprise Edition and is in the process of building new Oracle-based warehouses in North America and Europe as part of the corporate BI strategy of Henkel Consumer Adhesives' parent company, the Henkel Group. "Today, our design consists of star schemas built in to the relational database," says Rhoades. "A lot of the data-mining logic we code ourselves. I think that the data-mining application can automate a lot of that for us. And incorporating some of the new OLAP features will make querying the database even faster and easier for end users."

Richard Winter, president of Winter Corporation, a Waltham, Massachusetts, consultancy specializing in the technology and implementation of large databases, also sees performance benefits when the relational database, OLAP, and data-mining engines are integrated. "Historically, these engines have been developed by independent companies, and of those groups, only the database companies have had the time, resources, and engineering talent to fully develop the performance and in particular, the parallelism of their products," Winter says. With integration, the OLAP and data-mining engines benefit from the optimization, parallelism, and performance engineering built into the RDBMS. That will go a long way toward solving BI's traditional stumbling blocks.

"Integration is a huge problem for owners of large data warehouses," says Winter. "The whole challenge of building a data warehouse is that it's a huge integration problem. Gaining more efficiency in that integration process is terribly important to making the idea work."

He concedes, however, that the BI market may need some convincing that the all-in-one solution is the best option. "This is an alternative to the best-of-breed strategy, so it's like buying an audio system. You can choose each component according to the manufacturer that you like best or the model that you like best, or you could buy one system from one company."

Some customers voice similar concerns. Jim Barlow, senior manager for business intelligence for Georgia Pacific, is a BI veteran who doesn't take new technology lightly. His US$22.2-billion company, based in Atlanta, Georgia, relies on secure analytical systems to help meet current sales goals and scout out new strategies. The firm's largest database is a little under 1 terabyte of total data, and it's being used by 1,300 employees. So Barlow isn't one to adopt something new just because it's the latest and greatest innovation. "It makes all the sense in the world that you can get all your data warehouse and business intelligence needs in one package," Barlow says. "You only have to deal with one set of upgrades, one vendor. At the same time, we're very cautious of it because if everything comes together in one place and one thing fails, does that mean they all fail?"

Questions like these are natural and even helpful to vet new technology that offers a fundamental change in an existing architecture. But with BI capabilities embedded within Oracle9i, BI applications can take advantage of the same scalability and reliability that supports the core RDBMS. This includes support for Real Application Clusters, which offers users the ability to define highly available database services that are configured at the Oracle network level. The technology makes it easy to configure failover policies, and the "event notification of instance failures" feature provides fast notification for application servers and applications.

The demonstrable benefits of BI integration may quell the concerns of people charged with keeping important corporate systems up and running, according to Raymond Roccaforte, Oracle's vice president of data warehouse engineering. "The question will certainly come up, 'Should I continue to use my own [OLAP] server, or should I evaluate the new Oracle capabilities that are in Oracle9i?'," he says. "We've seen at a couple of our customers' sites that they don't like the stovepiping of data. They want better integration with the relational database. A lot of our customers intend to migrate up from Oracle8i to Oracle9i by the end of this year. All of a sudden, these customers are going to find that they have very sophisticated OLAP capabilities built right into the database. It's going to greatly simplify their environment," Roccaforte says.

Click on the image for a larger version.

Dismantling Stovepipes

To understand the potential of Oracle9i's integrated architecture, it's helpful to examine what's new in the architecture. RDBMSs historically ran as their own data kingdom, managing hundreds of gigabytes and even terabytes of data. To add analytics such as OLAP and data-mining queries to these systems, IT managers traditionally used a collection of commercial and homegrown technologies to construct data pipelines into discrete servers, like the venerable Oracle Express Server, the multidimensional database that built OLAP cubes out of relational tables. Once the data was extracted, transformed, and loaded into the Express Server, analysts had a powerful platform for manipulating key information in multiple views. Analytical applications, such as Oracle Reports and Discoverer, helped managers drill down into the data, but these worked only against the relational server—an example of the split between multidimensional and relational worlds.

Unfortunately, BI systems built with discrete servers, data stores, and analytical tools can be difficult to maintain and use. Instead of being pervasive, BI becomes relegated to individual divisions within a company that may come to rely on their own stovepipes of data rather than on centralized information that flows seamlessly throughout an enterprise. In unintegrated environments, BI is typically based on four separate processing engines: the relational database; an extract, transform, and load (ETL) engine; the OLAP engine; and a data-mining engine, each often with its own local storage in different formats.

"When you have to move the data into a separate multi-dimensional store, you generally have problems with translating the data, and you have performance problems in getting the data over there," says Dan Vlamis, president and CEO of Vlamis Software Solutions, a Liberty, Missouri, systems integrator and software developer specializing in data warehousing and OLAP. "There are many OLAP tools that will report the data directly out of the relational engine," Vlamis says. "What they lack is the ability to do more-complex, multidimensional calculations. What Oracle has finally done is let you do that analysis directly against the relational data."

Oracle9i Release 2 unifies all of these formerly separate engines within the RDBMS. When you only have to maintain one server, the result is BI simplification and cost savings. In addition, companies don't have to build and maintain the data pipelines that used to connect all the various servers when they were separate entities.

"We've been doing prototypes for customers who are migrating from standalone multidimensional databases to Oracle9i Release 2," says William Endress, Oracle's director of product management for OLAP technologies. "In one particular case, they will go from two servers and a total of 14 CPUs to a single server with four CPUs. That's a tremendous cost savings in hardware. And they'll no longer need two DBAs."

OLAP Benefits

To understand where the specific advantages are in integration, consider two key BI applications: OLAP and data mining. OLAP benefits from integration in five important ways:

A single copy of metadata. Integration ends the need for multiple copies of metadata. When BI systems use a separate RDBMS and OLAP engine, the latter requires its own metadata model. But to build the OLAP cube, a company must extract the data out of its data warehouse, transform the data, and map the new information into the cube. Companies duplicate a lot of the work that was already done to get the data into the data warehouse.

Dismantling data pipelines. The job of building efficient and scalable data pipes from the RDBMS to the OLAP engine can be time-consuming for technical staff. "You have to know the interfaces of both engines extremely well," Roccaforte says. "You have to know how to tune both engines and all the tricks of the trade for both engines." Integration eliminates these data pipelines.

Performance. Data analysis occurs faster within an integrated platform because there's less data movement between the RDBMS and OLAP servers. For example, the data transfer rates between relational datatypes and multidimensional data types in Oracle9i Release 2 are anywhere from 10 to 200 times faster than in a separate multidimensional database, according to Roccaforte. In benchmarks using real data, the basic operation of creating the OLAP cube and reusing the data later are an order of magnitude—or even two orders of magnitude—faster, he adds.

SQL for multidimensional data. The embedded OLAP engine gives analysts the ability to run SQL queries against the multidimensional database, which eliminates the need for proprietary interfaces and the limited number of analytical tools that traditionally access separate OLAP engines. Integration negates the need for data replication between relational and multidimensional data sources, when companies had to build a data warehouse twice: once for SQL-based tools working with relational tables and a second time for tools to go against the multidimensional database.

"Most of the query tools and decision support system (DSS) tools out there are SQL generators for relational databases," Roccaforte explains. With an integrated architecture, data analysts have a wider range of analytical programs to choose from, including ones they're already familiar with, when they need to run multidimensional queries. "You no longer have this technology divide that exists between relational tools and multidimensional tools," Endress says. "You can simply choose one toolset and use it against both relational and multidimensional data."

For Rockport, having multidimensional data that's accessible via SQL is "something that we've looked forward to for many years," says Bob Logan, managing director of Rockport Software. "There are so many SQL programmers around who have no knowledge of the kind of power that's in Express. Now they can finally use SQL to connect [to a multidimensional database]."

Reduced data-replication needs. BI users aren't required to replicate data between the multidimensional and relational data stores when they're combined into one database. The multidimensional data is backed up and restored during the normal maintenance of the relational data. "Thus, if you're a consumer package goods company, and you're starting an advertising campaign on Monday night, it's quite plausible that by Tuesday you'll start seeing the results coming through scanner data feeds from, say, grocery stores to see if your advertising campaign is effective," Endress explains. "But if you're building two different warehouses, the process could take three days, and you're not going to be able to adjust your advertising campaign until Thursday. However, if you can load it once into Oracle9i Release 2 and can begin analyzing that data immediately, you've now compressed two or three days out of the cycle, and you can adjust your promotional campaign in real time." hat does multidimensional analysis add to traditional queries against the relational database? First, the multidimensional engine specializes in complex multidimensional calculations that involve interrow calculations. Instead of asking the question, "Who are my top 10 customers?" analysts can ask something like, "What was my percent change in market share for a grouping of my top 20 percent of products based on sales for a current three-month period versus the same period a year ago for those accounts that grew more than 20 percent?"

"I actually see five calculations that occur and interact across dimensions within a single query," Endress says.

The queries might be based on hierarchical weighted averages. For example, an analyst might have a hierarchical weighted average of price as he or she aggregates data through a product dimension. "Obviously, if I just add all the prices in a price sheet and say the sum of all those prices equals the price of my brand, that doesn't quite make sense. You're probably going to do a weighted average based on a quantity to arrive at that figure," says Endress. The multidimensional engine could also provide weighted regressions, correlations, and deviations for financial calculations such as interest rates, depreciation, and payment schedules.

Data-Mining Integration

Similarly, data-mining operations gain a number of important benefits from working within a tightly integrated BI environment.

No matter what vertical market a data-mining user is in, the task is almost always the same: sorting through reams and reams of data trying to find important patterns that can unlock new business opportunities. To do this, many companies use their data-mining systems for information classification and prediction. "You're trying to take a look at the past behavior of people," says Charlie Berger, Oracle senior director of product management for data mining and personalization. A data miner might assign each person in the sample either a "0" or a "1" value—0s for loyal customers, 1s for churners. "You may sift through 700 fields of data from several hundred thousand customers to find patterns and profiles that identify the churners and nonchurners. Data mining helps you to find gold needles in the haystack," Berger says.

Until the tight data-mining integration offered by Oracle9i, companies had to extract representative samples of customer data from the secure and scalable relational database and move it into a separate data-mining server. The analyst then ran datamining algorithms on the data and built a predictive model. But to use the model to make predictions with the full storehouse of data, he or she had to again move even more data out of the RDBMS, run the model, and then move all the resulting predictions back into the database. "This requires moving the data around a lot," Berger says. The alternative is to take the model to the data—export the model as C, C++ or Java Code. "But then I have the task of wiring that application into the database to run that code."

By natively embedding Oracle data-mining capabilities and a Java application-programming interface (API) into Oracle9i Database, Oracle makes it easier for analysts to build applications that can uncover subtle behavioural patterns. The better place to perform all those transformations and data manipulations is inside the Oracle database.

"Because all the data is in there, you have all the power of SQL to do these transactions, and now it's just another step to build your predictive model. When you want to make your predictions on the other data, all your transformations are retained in the data. You make your predictions, and the results are still in the database," Berger says. "With Oracle9i Data Mining, I'm actually building a set of programs that I can just hook up to, say, Oracle Workflow or Oracle Reports. So once I build my predictive model, it scores them and produces a nice formatted report that sorts people according to their probability of churning or their probability of spending more than US$500 on their expected purchases," he says. "Now a mere mortal marketing analyst can sit down and say, 'I'd like to run that shoe campaign we did in New Jersey last quarter in New York, and I want to know who to send my e-mail or direct mail campaign to.' Well, he or she just has to pick the set of data that was used in the past and build a model on that data. I've already built an application that knows what to do with the data, how to apply the models, and where to put the results. We're enabling application developers with data-mining data analysts to build predictive business intelligence applications. It shortens the time to go from the raw data in the database to making predictions."

Thanks to the Java API, programmers have the option of building applications that automatically push these predictions out to the appropriate business people. For example, when a customer calls a telecommunications company with a service complaint, the data-mining application could send an alert to the service representative's computer screen that identifies the caller as a potentially highly profitable customer that shouldn't be lost.

Berger adds that because the data-mining server is within the RDBMS kernel, Oracle has been able to fine-tune the platform for performance. "When we need to go off and count things even faster or run in parallel, we're right there at the core of the database. These are all the types of things that we take advantage of by being in the database."

Performance numbers for the integrated datamining engine are comparable and in some cases faster than when the engine ran outside of the RDBMS, Berger says. In fact, some data-mining algorithms run as much as 10 times faster when run within the Oracle9i kernel versus on a standalone engine, according to Berger.

For a systems integration company such as Vlamis Software, the ability to create BI objects within the data-mining environment is an important strategic win, according to Chris Claterbos, a consulting and development manager for Vlamis. "Now we can go to a client and say, 'Hey, you already have Oracle, you already have these reports set up, and you're building a data warehouse. Why don't you put some business intelligence on top of that and get more value out of the data by doing some OLAP analysis against it?' The client may then say, 'This is just a little step. Let's look at that.' So for us, as integrators, this is huge," Claterbos says.

The Power of Integrated Reporting

Of course, all the analytic and mining functionality now integrated into the database kernel is useless if you can't leverage it with sophisticated reporting tools. That's where Oracle9iAS Discoverer and Oracle9i Reports come in. Oracle9i Reports allows users to create reports with data from any source and in any format. Business users of all skill and organizational levels—from casual users to experienced analysts—can access these predefined reports or use Oracle9iAS Discoverer to create and execute ad hoc queries.

Oracle9iAS Discoverer is integrated with Oracle Warehouse Builder metadata, so administrators can generate End-User Layers (EULs) from metadata in OWB. This provides the ability to drill down and see the source of the information behind a Discoverer query.

It was this advanced level of functionality that Georgia Pacific needed for its analysts. "It's a pretty sophisticated user base that we have, especially in the sales and financial and marketing end," says Georgia Pacific's Barlow. "They want to go outside the hierarchy and do a lot of 'what if' scenarios, and the Discoverer tool allows them to do that: to pull together multiple tables or hierarchies not normally put together to see the resulting data set."

Along with advanced functionality comes integration with other Oracle technology. Discoverer and Reports are tightly integrated with Oracle9i Database, meaning they can use the security, scalability, data-access, and analytical features of the database. And Discoverer is seamlessly integrated with Oracle9i Reports Developer, the enterprise reporting tool. This integration means that Discoverer users can extend their workbooks using Reports, and have access to all of the publishing services of Oracle9iAS Reports Services. With the tight integration between Reports, Discoverer, and Oracle9iAS Portal, users can publish workbooks and reports portlets or components of a Web page. ON Semiconductor, a Phoenix-based supplier of integrated circuits and standard semiconductors, leveraged the integration between Oracle9iAS Portal, Discoverer, Reports Services, and Forms Services to create an intelligent self-service reporting portal. An intranet portal with browser-based Web analytic tools allows the business users to quickly access and derive information from large amounts of data stored in an Oracle data warehouse—all from a single, integrated, Web-based desktop environment. According to Jim Hill, manager of data warehouse operations for the company, tight integration and centralized data management mean that all users have access to consistent, up-to-date data. "One of the benefits that we see with Oracle is we can architect a business logic into the end-user layer," says Hill, "and when we change the way we do business, we make those changes in the end-user layer and it ripples through all of our reporting."

Coping with Complexity

As the needs for up-to-date, real-time information grow, so too do demands for speed, flexibility, and easy maintenance of the underlying BI system. In the past, cost and technical complexity were enough to limit BI applications to resource-rich IT companies and departments. The collection of technologies now integrated within Oracle9i Database go a long way toward answering the price, performance, maintenance, and data-quality challenges of BI.

Alan Joch (ajoch@monad.net), a New England-based technology writer specializing in enterprise and internet applications, is the author of How to Find Money Online: An Internet Guide for Entrepreneurs (McGraw-Hill, 2001)

Highlights: What's New in Release 2

By Alan Joch

New in Oracle9i OLAP

Simplified management. Specialized administration tools necessary for a separate multidimensional database are eliminated.

Improved security. One security model handles both relational and multidimensional data.

Easier and faster analyses. Both the relational and multidimensional data are accessible through SQL and OLAP application-programming interface (API) clients.

Composite range-list partitioning for rolling window operations on a list of partitions. List partitioning supports default partitions; parallel DML is supported on nonpartitioned tables.

New in Oracle9i Data Mining

Automatic recommendations of the most accurate data-mining algorithms to run.

Assistance finding subsets of variables or fields most predictive for a target field.

Two new types of clustering for finding naturally occurring record groupings in the database.

New in Oracle9iAS

Centralized management through Enterprise Manager and support for Oracle Single Sign-On.

Tight integration in Oracle9iAS Portal means Oracle9iAS Discoverer workbooks and Oracle9i Reports can be published as portlets with easy customization and automatic refresh features.

Oracle9iAS provides access to new Oracle9i analytic functions and the ability to create data cubes for enhanced aggregation and "what if" analysis.

Oracle9i Reports supports data from any source, allowing development of enterprise reports in most formats with distribution to any destination.

Oracle9iAS Personalization delivers real-time recommendations to customers using Oracle9i Data Mining technology.

Oracle9iAS Clickstream Intelligence provides insight on Web site effectiveness through 150 predefined reports based on Web server logs.

Oracle9i BI Beans allows developers to customize applications using Oracle9i OLAP services.

Alan Joch (ajoch@monad.net), a New England-based technology writer specializing in enterprise and internet applications, is the author of How to Find Money Online: An Internet Guide for Entrepreneurs (McGraw-Hill, 2001).

Taking BI Personally

By Alan Joch

By tying the latest release of Oracle9i Application Server together with the Oracle9i Database Release 2, developers of pioneering Web applications are discovering the benefits of data mining and real-time recommendations.

Newly incorporated into Oracle9iAS is Oracle Personalization (OP), technology that can unlock important customer information for Web businesses or provide the foundation for sophisticated custom-marketing applications.

OP can analyze customer demographics, purchase histories, and data about Web site navigation to suggest new products and services to online shoppers. OP sorts through all that data automatically and can suggest a new book or video that matches a customer's profile, or it can suggest a banner ad targeted to the shopper's buying patterns.

Swisscom Corporate Technology, the R&D arm of one of Switzerland's largest telecommunications providers, is using OP to prototype a unique application that directs moviegoers to preferred films. The Bern-based unit of Swisscom recently finalized the prototype for an application called MovieGuide, which not only tells people what movies they might like to see, but it also uses global positioning technology to direct them to where the movie is showing. And unlike newspaper reviews or the recommendations of friends, MovieGuide bases its advice on the tastes of the individual moviegoer.

When somebody registers for the MovieGuide, the film buff is asked to rate at least ten movies he or she has seen in the past. MovieGuide users also rate new films as they see them, and in the process the system refines its knowledge of each user's likes and dislikes.

Behind the scenes, OP matches new movies to what a subscriber liked in the past or makes recommendations based on the preferences of moviegoers with similar tastes. "This is all based on data mining," explains Christoffer Swanström, project leader at Swisscom. "You're correlating the tastes of different users."

In the current prototype, MovieGuide users can receive suggestions via a PC or any PDA that runs Windows CE. The guide also supports mobile phones with SMS access, so users can type in the keyword and receive text messages with recommendations. Swisscom plans to add a WAP interface in the near future.

Creating the application was relatively easy from a programming standpoint, Swanström says. "In OP, we call the Java API [application-programming interface] from our application. We have a number of servlets behind the MovieGuide that make these calls. Behind the API, the system is querying the database directly. But what we see as developers is really just a Java API."

Similarly, GIP AG, a Mainz, Germany, data-mining consulting company, used OP to develop a prototype for a Web game site. The application watches gamers as they answer trivia questions about a variety of subjects and then provides questions based on each player's areas of interest. Better questions increase player interest and result in "stickier" participants, says GIP researcher Thilo Simon. "If a player is a science fiction fan, then you can ask him more or fewer science fiction questions to make the game easier or more difficult," Simon explains. "This is happening in real time."

Because of OP's Java API, GIP didn't have to change the production database that's associated with the application. "OP is a second, parallel system that stands beside your production system. It only makes the recommendations when you ask for them. This is all happening automatically as the player is playing the game."

The only prep work required is to tell the system when a new player registers and whether or not players answer each question correctly. Based on that input, the application automatically updates the underlying data-mining models each night so it can offer new recommendations. "You can tell OP to take this collected data and build a new model, say, every day, every week, or every month," says Simon. "As you gather data, you get more-sophisticated models so you ask better questions." And asking better questions is what BI is all about.

Alan Joch (ajoch@monad.net), a New England-based technology writer specializing in enterprise and internet applications, is the author of How to Find Money Online: An Internet Guide for Entrepreneurs (McGraw-Hill, 2001).

BI Beans Speed BI Application Development

By Alan Joch

The Oracle9i platform provides a simplified development environment that speeds creation of BI applications. With high-level, Java-based building blocks, BI users can quickly assemble BI applications that sit on top of Oracle9i.

At the core of this environment are JDeveloper and BI Beans. As of Oracle9i Release 2, JDeveloper includes wizards for creating graphical crosstabs, graphs, and other elements for analysis and presentation, all without forcing end users to write any code. Instead, application generation wizards step users through the series of programming tasks needed to create Java-based, thin-client applications. "BI Beans offers a JavaServer Pages (JSP) tag library that allows you to insert custom tags inside a JSP page," explains Marty Gubar, Oracle's director of product management for analytic tools. "The custom tags provide rich, out-of-the-box, analytic functionality for end users. For example, a single tab enables a user to display data as a crosstab or graph, slice and dice the data, and drill down from summary-level data into details."

Similarly, developers can create thin-client servlets and Java client applications using wizards that will generate application code.

Gubar says a major trend in ERP and CRM applications is to embed business intelligence into the applications without requiring end users to use a separate tool for business analysis. For example, a CRM application may enable a salesperson to analyze the sales activity for a customer that he is about to call. "We're seeing what some people call business intelligence for the masses: having analytics as part of an existing application," Gubar says. "Oracle9i JDeveloper with BI Beans has made it very easy to take transaction applications and add business intelligence."

The JDeveloper platform simplifies life for BI consultants, according to Chris Claterbos, consulting manager for Vlamis Software Solutions of Liberty, Missouri, a systems integrator and software developer specializing in data warehousing and OLAP. In the past, Vlamis first had to decide if an application it was developing would be deployed in a client/server environment or on the Web. Each environment required different tool sets and different development skills. Now, with a Java-based platform, programmers just tap into their Java development skills, and they don't have to make any up-front decision about where the application will run. "You can choose that at the time you deploy, rather than at the time you develop," Claterbos says.

The JDeveloper wizards provide a rapid-programming framework that lets developers complete applications in a day or in three months, as with large projects such as Vlamis's VSS Business Analyzer, a commercial reporting and analysis tool designed to run against Oracle9i OLAP. "VSS Business Analyzer would normally have taken a development team a good six to nine months to develop," says Dan Vlamis, president and CEO.

Alan Joch (ajoch@monad.net), a New England-based technology writer specializing in enterprise and internet applications, is the author of How to Find Money Online: An Internet Guide for Entrepreneurs (McGraw-Hill, 2001).

© Rockport Software Limited, 1998 - 2009
All product and company names and logos are used for identification purposes only, and may be trademarks of their respective owners

• Home
• About Us
• Oracle BI
• Software
• What We Do
• Careers
• Links
• Contact Us
• Sitemap