Fedsure Life, South Africa`s fifth largest life insurer, is engaged in enterprise-wide business process re-engineering (BPR). In this process it is moving its focus from host-based processing to client-server systems and deploying a new line-of-business administration system, running against DB2. To speed implementation of the new system, to capture knowledge of the new processes, to enforce standards and consistency, and to guarantee a high quality of application, Fedsure Life is using the Cayenne data modelling toolset, licensed by Software Futures. "It`s difficult to design large, complex systems without modelling upfront," says Hennie Reyneke, life business manager, Fedsure Life. "Data modelling ensures referential integrity, it helps you define the relationship between entities and automates implementation. It also makes sure we stick to standards and have consistent screen displays across applications." Fedsure Life`s existing system was a solid performer, but policy-based rather than client-based, says Reyneke. "This made us inward-focused. So we`re changing our systems to reduce the time take to open a policy, settle a claim or bring new products to market." Modelling with Cayenne has many benefits, says Reyneke. "It`s a formal way of working. When you use models, the business owns the systems. This helps bring business and IT closer together. And if the business wants to change, data modelling makes it relatively easy for business and IT to work together in joint application development sessions. "Because we`re putting in so much of the effort upfront, maintenance should be reduced. And at the end of the day we, rather than business consultants, own the view of our systems. Finally, data modelling accelerates the induction process and makes it easy to bring new system developers up to speed." Fedsure Life next intends to model other aspects of its business: point of sale, workflow and image, and executive information systems; and to model the relationships between the various entities. To put Cayenne`s data modelling into perspective, Reyneke says: "Had we applied Cayenne before creating our existing system, our Year 2000 problem would have been relatively easy to solve."
Share
Editorial contacts