Building Intelligent Agents Using Business Activity Monitoring

Information Management Magazine, December 2005

Mike Ferguson

Most companies today have a significant investment in business intelligence (BI) to measure business performance. To date, most of these BI systems have been built to support decision making in specific business functional areas such as marketing, finance, HR and so on as opposed to being built around core operational processes. Most businesses are probably not using BI to continually and automatically monitor events in their operational business processes as their businesses operate to rapidly respond to detected problems or to predict if problems lie ahead. In general, therefore, companies have no active real-time element to their BI systems. The consequences are that nothing is helping the business to automatically respond immediately when problems occur or opportunities arise. Also, there is no automatic notification or flagging of alerts to take action that may avoid unnecessary costs, business disruption, operational mistakes and unhappy customers in the future.

Yet, the business processes that span all the functional areas, from front-office order entry to back-office shipping, billing and payment collection, govern how businesses operate. Making sure that core processes perform optimally has a major impact on overall business performance.

This lack of BI systems involvement in everyday operations often means that business executives see BI as passive, with no active role in managing operational performance to optimize real-time operations. To implement operational performance management, companies need to extend their BI systems, adding new technology components to help integrate BI with operational business processes. These components can be used to automatically guide people and optimize operations. Examples of such automation include:

• Automatic monitoring of events in operational business process activities as processes execute to see if any immediate action needs to be taken either inside or outside the enterprise.
• Automatic, rule-based action taking, whereby technology can continually test for specific business conditions and changes in operations. If necessary, alert people, make recommendations, send messages to other applications and/or invoke whole business processes when such conditions or changes do or do not occur.
• Automatic alerting upon detection of an actual business problem or opportunity, e.g., a canceled cargo order will leave a ship half empty; immediately alert sales to resell the new capacity.
• Automatic alerting to draw attention to a specific predicted business problem or predicted opportunity, e.g., it is predicted that an aircraft engine part may fail the next 20 flights, therefore just-in-time maintenance needs to be scheduled on the next landing while the aircraft is in turnaround. Alternatively, an order change requires an increase in product manufactured and additional distribution capacity to manage the additional product(s) ordered.
• Automatic action messages sent to other applications and processes to cause action in response to detected problems and opportunities.

One of the key types of event-driven BI processing that makes these kinds of automation possible is business activity monitoring (BAM).

What is Business Activity Monitoring?

Business activity monitoring is the ability to automatically monitor events associated with specific activities in an executing business process. Some would argue that BAM can monitor events whether or not they are process related. That may well be the case, but it is difficult to put events into a business context if they are not associated with a business process. Given that process execution is managed by business process management software, BAM technology can be very effective when it is integrated with process management software that is governing the execution of processes. BAM software integrated with process management software allows us to extend BI systems to monitor process activities.

Figure 1 shows an example of an order entry, fulfillment and tracking process, consisting of several process activities. BAM allows companies to select specific events to monitor that are associated with specific activities in a business process. Note that BAM can be used to monitor executing processes in multiple places. Using the example process in Figure 1, a company may choose to monitor events associated with orders, package assembly, shipping and payments. This means that multiple BAM jobs or intelligent agents can be created to monitor different events in different parts of a business process as it executes. Multiple intelligent agents can be built and deployed across business operations.

Figure 1: Monitoring Events Associated with Specific Process Activities

Possible business reasons might be:

• To detect any changes in orders (canceled orders, large orders, orders from valuable customers) and, if need be, to alert the sales force and other people in operations when specific events have a business impact, e.g., to respond when canceled orders occur.
• To check orders or predicted demand against inventory to optimize a supply chain by reordering/delaying/canceling supply of inventory.
• To detect or predict bottlenecks in the package assembly activity.
• To detect or predict delays in shipments for valuable customers.
• To detect or predict cash flow problems because of late payments.

In order to implement BAM and create multiple intelligent agents, companies need technology to allow them to:

• Listen for and capture specific event(s) as a business process executes.
• Integrate a necessary set of data using event-driven data integration technology as soon as each specific event being monitored occurs and to make this data available for automatic analysis. This data may come from data warehouses, BI reports, cubes, operational systems, external data sources or some combination thereof. Note that the target is not a database. This data is handed directly to a model for automatic analysis, i.e., the target is a service.
• Automatically analyze the selected integrated set of data to produce intelligence.
• Use rules to automatically determine whether to take action on the intelligence and if necessary to actually take those actions.

Page 1 of 3.