Embarking On A Service Oriented Architecture - The Intelligent Way

As organisations seek greater growth and the ability to adapt more quickly to change, there is a greater need for insight into what impact business processes have in terms of business performance and the ability to achieve organisational goals.

Recently, there has been a significant growth in demand for pervasive business intelligence (BI) to gain better visibility into business process performance and improve decision making at all levels of an organisation. According to IDC, the Asia Pacific (excluding Japan) BI and Financial Performance and Strategy Management Application markets are estimated to reach $724 million by 2010.

At the same time, organisations are also looking to Service-Oriented Architectures (SOA) to better leverage their existing IT application assets to adapt more quickly to change and seek greater growth. IDC anticipates SOA adoption in Asia Pacific to grow as businesses look to align IT with business processes. CIOs and IT managers have identified building and acquiring new applications, integrating applications and consolidating hardware infrastructure as their top priorities, whereas end-users have expressed that SOA adoption will help with better and more transparent business processes alongside more effective control and monitoring.

The simultaneous growth of investment in BI and SOA technologies is no coincidence. Today, SOA and BI have emerged as top strategic focus areas for CIOs. While both are able to bring considerable benefits to the organisation when deployed separately, there are significant synergies which can be attained by integrating BI within a SOA.

Areas that organisations looking to take advantage of the synergies of integrating business intelligence within a SOA should consider are:

Understanding business process performance

Being able to analyse the performance impact of business processes executing in a business process platform is a key objective for organisations developing SOAs. Once the business process data is captured in a database, data warehouse, or even flat files, it can be accessed by BI tools and further integrated with data from other sources, with the dashboards used to present aggregate views of the information. This helps provide further insight, allowing organisations to determine what changes should be made to a business process to improve business performance.

For example, orchestrating the delivery of a cell phone to a customer may generate data such as time spent provisioning the cell phone, updating the appropriate accounts and delivering the product to the customer. The BI tool can combine this data with data stored from other systems, such as usage data stored in another relational database by a billing system and business intelligence may be used to present an aggregate view. By combining call usage data with process cycle time data captured from business intelligence tools, organisations can gain insight into the impact of process delays, such as the amount of lost revenue that can be attributed to delays in the process.

Greater business agility requires business insight and visibility. If businesses have no insight or visibility into the impact a business process has on business goals, then it would be difficult to determine what changes should be made to a business process to improve business performance.

Action from insight

When a business user gains insights that flag a business performance problem, some kind of action will typically need to be taken in order to address the problem. Often, the action may involve invoking a business process. If this is difficult to achieve, then there is often less value to the insight.

While basic built-in application-level monitoring is good at triggering events or raising flags for individual applications, there is usually no way to correlate these events to events coming in from different applications or the affected business processes. Business activity monitoring (BAM) is a key technology component that allows business users to build real-time operational dashboards, thereby providing them with the ability to monitor business processes to understand the impact on key performance indicators (KPIs) affecting their business to improve operational visibility.

Integrating business processes with BI tools enable business users to take appropriate action in response to the insight they gain through interactive dashboards, reports and alerts. For example, an interactive dashboard can display a set of KPIs that are used to monitor given areas of business performance, with red lights against KPIs that are not met, such as product revenue drive being below target. With business intelligence tools, scorecard menu options on the dashboard can provide relevant actions to further investigate by navigating to more detailed dashboards.

Thus, business users can then directly take action as a result of the insight they gain from the business intelligence dashboard. Business processes can also be invoked as a result of an alert.

Orienting business processes to business goals

A common business requirement is to define business processes that drive the organisation towards key business goals defined in terms of KPIs and business metrics. These KPIs often derive from information scattered in many data sources throughout the organisation.

In traditional manual business process implementations, BI tools are used to make better decisions based on good business insight and quality data. For example, the selection of a supplier may be based on key business goals for driving down costs or improving customer satisfaction. Thus, a supplier may be selected based on performance over a given period with respect to price, reliability or responsiveness. This kind of supplier performance information is frequently accessed by a person using business intelligence tools and enables decisions to be made that drive an organisation towards key business goals.

The integration of BI with business process platforms in this manner enables the use of business intelligence to make analytic decisions in the context of automated business processes as well as providing access to key business data the process needs.

Considerations when integrating BI within an SOA

When integrating BI within an SOA, organisations need to consider the scale of the underlying business processes and understand how it may impact the business service levels. For example, a lead-to-cash process for a large company may have tens of thousands of concurrent instances, thus poor scalability or availability of the business intelligence services can severely impact the business.

In addition, performance of queries invoked by business processes need to be appropriate for the level of responsiveness demanded by the process. The use of data warehousing technology and best practices to integrate business process data can also substantially improve the query performance.

Another consideration is the degree of coupling between the business process platform and the BI system. In particular, it is important to ensure that business processes are loosely coupled from the physical data sources and the semantic metadata definitions of calculations, metrics, hierarchies and dimensions.

Conclusion

Greater business agility requires greater business insight and vice versa. Integrating BI within SOA provides an essential synergy. Further, the ability to access BI via web services interfaces enables sophisticated analytic and reporting capabilities to be accessed from within automated business processes. Users can also take action from the business insight gained by invoking business processes directly from BI dashboards.

Singhal is director, Fusion Middleware Sales Consulting, Oracle India.