| Popis: |
The application of agile software development is primarily an art. Organizations are adopting agile software development methodologies through a combination of bottom-up adoption and top-down change. However, reality of agile adoption has diverged from the original ideas described in the Agile Manifesto, with many actual adoptions resembling what is labeled as Water-Scrum-Fall by Forrester Research. In large organizations, development teams have adopted some variant of agile, for their product development efforts, what happens before and after this effort is still very traditional and waterfall.Not only is the performance of successive iterations difficult to compare, the results of agile adoption case studies are also difficult to compare as they vary in the scope of application of agile methods (most use only a subset of practices). Thus this thesis follows an exploratory case study approach to document the root cause for the water-scrum-fall phenomenon at a large scale Enterprise and presents the observations of a new initiative at the Enterprise to expand the scope of Agile development. Further, the aims of this case study thesis are to understand the limitations of Water-Scrum-Fall, and push back on the Water-Scrum side of the model. To do this, we model the as-is process (ie. the traditional handoffs model) and the to-be process (ie. Iterative requirements specification process) within an overall predominantly waterfall-based delivery model. The to-be initiative is still in its infancy, yet it provides new insights that we capture by a precise workflow for Iterative Solution Scoping to Deployment (ISSD) model. Through the use of the ISSD model on a pilot project being implemented at the Enterprise we identify lessons learned thus far. Specifically through the means of the ISSD, we show that the expanded scope facilitates: 1) regular feedback into the decision process for Continuous Improvement of Iterative Solution Scoping process and illustrate its performance-based utility, 2) better negotiations through interactions between the client and the development team by increasing collaboration, 3) define precise points of metrics for data collection which helps in better understanding the sources of performance deficiency towards managing variability, 4)change of mindset in letting go of the waterfall-related idea of "planning to commit," where development teams plan as much as possible upfront before committing to those plans, and instead shift to "commit to planning", where planning is a constant process throughout the development lifecycle, and 5) frequent and coordinated release cycle that results in a higher business value delivery and reduced risk. |
