Organizations have two universes in constant flux; business architectures and IT architectures. Now a third factor has entered the mix – services oriented architecture (SOA). As organizations seek to align business and IT architectures, SOA can play a key role in streamlining this process. This article discusses how SOA helps align business and IT architectures to deliver more effective, more efficient responses to ongoing business demands – on a transitional basis and over the long-term.
Leveraging Services in Architecture Alignment
Business architectures shift on a continuing basis. Consider efforts to align business units, personnel and processes driven by mergers, acquisitions and new market opportunities. The consolidation, streamlining and automation of business processes also drive shifts in business architectures. Finally, day-to-day tactical demands and cost containment directives drive continuous changes across user requirements.
Opportunities abound if businesses can synchronize as-is and evolving business processes under a services oriented architecture. Some organizations have created a business framework called a “sparse matrix” into which analyst teams populate processes as they are analyzed and refined. The sparse matrix is a course grained business process and organizational model.
Populating this matrix requires understanding as-is business architectures, incorporating detailed business processes and mapping these processes to back-end applications. Top-down, as-is mapping of the business architecture also has the added advantage of creating a framework for SOA based deployment across frontline user environments. Front-end, SOA based architecture evolution enables and is enabled by phased alignment of business processes, back-end applications and enterprise data architecture.
Front-end Architecture Evolution Enables Business / IT Alignment
Front-end architecture evolution enables SOA because frontline users are the ones that understand detailed business processes within a given area. While the sparse matrix provides a high-level, cross-functional business view, filling in the details requires input from the trenches. This bottom-up approach is not merely a passive analysis of processes but a proactive initiative to automate, integrate and streamline frontline processes based on an incremental ROI model.
Using this approach to fill in the sparse matrix must be coupled with value-driven projects that automate and streamline business processes for frontline users in various business areas. Opportunities abound in this area. Business units have created application and data architecture workarounds in the form of spreadsheets, email, faxes, paper and pencil, and other user-based systems. These areas represent the front lines of process automation, architecture alignment and SOA deployment.
Business and IT analysts, arranged in small, rapid response teams, must work with frontline users to automate, streamline, refine and integrate processes leveraging leading edge technologies. Automating workaround functions serves a multitude of purposes. First and foremost is the fact that automation of workaround processes expedites customer response times; streamlines operational efficiencies; allows management to reallocate user resources; exposes hidden processes; and creates a foundation for subsequent automation, consolidation and integration.
Rapid response solutions must be SOA based by nature. Deploying reusable services under a common SOA framework creates flexibility and agility for all subsequent frontline deployment efforts. As frontline systems expand, they may be integrated with related frontline applications – establishing the foundation for emerging enterprise architecture.
One key aspect of frontline user-driven SOA systems is that they must be integrated with and across backend systems, eventually melding into a common architecture. Initially this will involve middleware and the integration of redundant back-end applications and interfaces using intelligent front-end applications. This increases ROI opportunities in process automation and integration while identifying back-end modularization and SOA migration options. Over the longer term, back-end functionality and frontline architectures merge into the future enterprise architecture.
IT Architecture Realignment and SOA
Creating new services from scratch to replace operational application environments is unrealistic given the substantial foundation of software assets, data and business logic embedded in existing IT architectures. One essential aspect of an SOA strategy, therefore, involves the incremental transformation of existing software assets into SOA based assets.
This practical approach to SOA evolution works effectively when coupled with the evolution of the business-driven SOA emerging at the frontlines of the business. As the sparse matrix evolves along with peripheral systems that automate, streamline and integrate business processes and existing front-end systems, back-end SOA migration options will begin to emerge.
Consider, for example, a front-end set of systems that automate and integrate a series of billing processes, including front-end interfaces to three redundant, mainframe billing applications. The next logical step, given that these front-end applications have already automated and consolidated business processes across billing units and regions, is to build and deploy a phased extraction and migration strategy.
In this scenario, analysts can begin to decouple back-end functionality through a process of modularization, extraction, componentization and decommissioning of mainframe logic. Modernization tools and techniques support this approach. As front-end, SOA based architectures are populated with back-end functionality, analysts will need to address data architecture issues.
In our billing example, this would involve rationalizing business data and system data usage across billing units and applications. This would be followed by the normalization of data, creating a new relational data model. This supports alignment of the billing business architecture and data architecture. SOA transformation, in conjunction with data architecture transformation, is supported by isolation and modularization of business logic, user access logic and data access logic under the SOA.
The Self-Aligning Business / IT Architecture
The above approach melds the concepts of organizational alignment; business process automation and integration; data and business architecture alignment; and phased SOA migration under a comprehensive business / IT alignment strategy. This approach sets the stage for a phased evolution to SOA, but more importantly creates an environment in which business / IT alignment can evolve naturally as the business changes.
This is based on the fact that agile front-end applications have been deployed using SOA based concepts and concurrently aligned with streamlined business processes as part of the design process. This coupled with the fact that business and data architecture alignment was incorporated into back-end systems modularization creates an environment where process changes can be rapidly deployed through the use of services within this new architecture.
