A Fortune 1000 Insurance Company
As with most software, the consumer is subject to updates, new versions and eventually deprecation. The latter was the case for the version of IBM Business Process Manager our client was using; thus, the client needed to upgrade to the latest version: IBM Business Automation Workflow. The new version, with new features and functionality, provided the client a great opportunity to modernize along with the upgrade. It would seem silly to go through the pains of an upgrade and not take advantage of the new capabilities.
However, with approximately 250 process applications to migrate, and with a dire need for modernization, it wasn’t going to be a simple operation. So, we conducted a thorough discovery session with the client to find the best approach, as we knew simply throwing bodies at the problem wasn’t going to cover it. We agreed the “work smarter, not harder” approach would be best for this effort.
Now that we agree on the “Work Smarter” approach, it is time to look for the areas we know we can save our client’s money. Salient Process has developed some industry-leading toolkits, templates and accelerators that we can leverage to help automate some areas of development. If development time per application can be reduced by 10 hours, we would in this case, be saving approximately 2500 development hours which we use to provide value in other areas, such as modernization of applications, training in best practices or new functionality. In this case we knew we would not be migrating 250 process applications overnight, so we worked with the client on an iterative strategy. So where do we start?
As is the case with any project focused on increased reuse and modularity, taking inventory is a critical first step. Seven services that all do the same thing (or slight variations), for example, is in direct conflict with the modus operandi, and identifying all seven is the first step towards consolidation. This exercise prevented us from “reinventing the wheel”. The negative flipside of this is “paralysis by analysis”. By assembling the right team and relying on Salient Process’s extensive experience, we were able to chart a course after only a few weeks of due diligence.
Since we can’t migrate and modernize 250 processes overnight, we also had to formulate a strategy for rolling them out incrementally. We considered several options, but the one that made the most sense to the client was by business areas. This kept the communication more focused during the delivery periods. It also resulted in more narrow deployments for end users. This proved to be a very manageable, yet efficient, approach for the client.
In terms of reusability, we found some patterns that were prevalent across many (if not all) of their applications. This was great news as it meant we could begin to harvest assets. In the Lombardi coaches (now deprecated), this had not been possible. Meaning that if something was used in 500 places and it needed to change, 500 individual updates would be required. By using coach views, reusable services and toolkits, we were able to not only cut development to a fraction of the original time, Dut the output was also a lot easier to maintain.
One reusable component of note was our Template coach view. This was a coach view that wrapped every client-side human service. It offered the developer a suite of reusable services including common Dutton events and validation and error handling frameworks — all as configuration options.
This Template accomplished several things:
Between the Template and various other reusable components, we were able to not only decrease the time it took to modernize in half, but also increase the developer count by lowering the barriers to entry for less technical resources. When you look at that cost savings, the two weeks of due diligence seems more than worth it.
Continuous improvement is a continuous process. Salient Process is still actively working with this client to progress the modernization initiative. Since the start of this project, we have rolled out three lines of business (approximately 30 process applications with another 30 expected by the end of next month). Bug fixes have been minimal since so much of the functionality is shared. And, although our job is far from done, we have already been able to accomplish something far more effective and far less cumbersome than simply throwing bodies at it.
Contact Us
