In-Depth

Making Real-Time Business Decisions with BAM

Business activity monitoring pure-play Celequest says that it goes where conventional BI vendors fear to tread—into the heart of real-time BAM.

• By Stephen Swoyer
• 07/02/2003

If business intelligence (BI) and enterprise application integration (EAI) players such as Informatica Corp., WebMethods Inc., and Tibco Software Inc. (among others) are to be believed, something better has come along, as what research firm Gartner Inc. calls business activity monitoring (BAM) (http://info.101com.com/default.asp?id=1134).

These vendors and others have formed partnerships, made acquisitions, or announced initiatives meant to spur the development of a technology that research firm Gartner Inc. says will be one of the top four initiatives driving IT investment and strategy by 2004.

To listen to BI-stalwart-cum-BAM-player Informatica tell it, the market for BAM software and services will be dominated by large vendors such as his company and its partner, WebMethods. “Unlike the niche vendors or the folks who are trying to build out this stuff and add to their existing platforms, because we have the best of breed technologies in ETL, EAI, and BI, we can create this infrastructure for the future as well as satisfying their BAM needs.”

Not so fast, says Celequest Corp., a small BAM start-up with a distinguished pedigree. It argues that the market for BAM solutions won’t be the exclusive provenance of the big BI and EAI players.

Celequest, which unveiled its Celequest 2.0 BAM software stack this week, is one of a cadre of several dozen pure-play vendors going up against the big guys in the still-emerging BAM space. If all BAM pure-plays have as compelling a story as Redwood Shores, Calif.-based Celequest—which boasts a non-disruptive BAM architecture that interoperates with existing BI and EAI software—there could be plenty of room for everybody.

Celequest is headed by Informatica co-founder Diaz Nesamoney. While with Informatica, Nesamoney served as both its president and chief operating officer. During that time, the data integration specialist transformed itself from a strict purveyor of extraction, transformation, and loading (ETL) tools to become a BI player with BAM aspirations. Nesamoney left Informatica in July of 2002, launching a start-up company called ViewCeler in September of that year. By June 2003, with the unveiling of its flagship product close at hand, ViewCeler changed its name to Celequest.

Celequest's Nesamoney argues that because of the traditional emphasis on batch processing, most BI players are ill equipped to respond to the requirements of a BAM marketplace. “Operational BI is mostly about delivering reports, but with BAM, what we’re really trying to do is help [business executives] make decisions in real-time,” he comments. “If you look at what’s happening to most BI companies, the larger BI companies, they’re trying to take their existing BI infrastructure, whether it’s an ETL tool or whatever, and they’re evolving it to become more event-oriented. But it’s still based on a batch-oriented architecture that was designed for data warehouse-style BI and reporting.”

In place of operational BI, in which data is commonly moved to a staging area—an operational data store—prior to being fed to a data warehouse, Nesamoney describes a memory-oriented BAM architecture—a physical memory cache – that allows an organization to more easily make changes to a data model. That’s because in a data warehouse, Nesamoney argues, organizations are constrained by the limitations of physically stored data. “With a data warehouse, it involves dropping in data and then indexing it, which could take some time. Secondly, storing data before you report on it reduces quite a bit of latency, even if you’re trickle-feeding it.”

Celequest 2.0 comprises four components: an Activity Server, a Scenario Modeler, an Activity Dashboard, and an Application Developer Workbench.

Celequest Activity Server provides a staging area for data, events, and messages, and includes what Celequest terms a Streaming Data Flow Engine—or a “Streaming Data Store”—which is a refinement of an operational data store (ODS). The Activity Server also materializes and caches real-time temporal views of events and integrates them in the context of historical data. It exploits a Adaptive Modeling Engine that executes business rules, manages exception conditions, and enables dynamic event modeling, including time-series analysis.

Celequest Scenario Modeler lets business users model analytical scenarios dynamically and evaluate them in real time. As its name describes, Celequest Activity Dashboard gives business users visibility into real-time metrics and alerts and enables users to acknowledge and initiate actions. Finally, Celequest Application Developer Workbench is a tool that allows developers to create baseline business views and data models that provide a real-time representation of business activity.

Nesamoney stresses that Celequest 2.0 is designed to complement existing BI infrastructures. It will support any of a variety of JDBC-compliant analytic dashboards, for example, and can use JDBC to query relational databases. Celequest provides agents that plug into packaged ERP suites from SAP, Siebel, PeopleSoft, and other vendors to detect and forward events to Activity Server. For integration with EAI vendors such as WebMethods, TibCo, and others, Celequest 2.0 can exploit Web services interfaces or vendor-supported APIs to detect and forward events.

Celequest is not designed to replace a customer’s existing data warehouse either. To provide historical data in context, for example, it requires that data, events, and messages be offloaded to an ODS or other repository from the Activity Server. Not surprisingly, Nesamoney says, many customers already have data warehouses that provide such information, “so we can fit into that environment pretty easy without disrupting what they’re already doing.”

If Nesamoney demonstrated prescience when he helped to co-found Informatica in 1996, he could be ahead of the curve this time as well. Although BI and EAI vendors have pounced on the nascent BAM market, most have found that they don’t themselves provide the necessary solutions to effectively capitalize on it. So they’ve looked elsewhere to flesh out their BAM portfolios. Informatica and WebMethods, for example, recently expanded upon a partnership first announced last summer; the two companies are now jointly marketing a combined BAM solution. EAI specialist Tibco gave the BAM pure-play market a visible boost when it acquired pure-player Praja Inc. last year.

The upshot, Nesamoney argues, is that there’s a unique opportunity for BAM pure-play vendors, which aren’t trying to adapt an existing architecture in to a new technology model. “Customers should consider pure-play vendors such as ourselves, just because we have the innovation and the technology to solve these problems,” he concludes. “My belief is that we’ve taken the right architectural approach and probably solved this better than existing BI vendors.”

There may be something to his argument, analysts say. In a recent report, for example, Gartner principal analyst Bill Gassman suggested that “the primary source of innovation in BAM [is] coming from pure-play BAM vendors.”