Architectural Decision Modeling Add-In for Sparx Enterprise Architect
Context matters when it comes to experience sharing; therefore simplistic best practices rules and design-by-authority are bound to fail in the real world. This makes architectural decision knowledge particularly precious. However, such knowledge changes frequently, and architecture documentation budgets typically are very limited. Therefore, knowledge reuse by chance is not going to happen.
Decision guidance models, created with the ADMentor tool, fill the gap between static and stale reference architectures and patterns and retrospective decision capturing in meeting protocols, project wikis, or software architecture documents.
- Problem space modeling: recurring design decisions, options to be considered - providing a checklist effect
- Solution space modeling: decisions made and their rationale (as mandated by the ISO/IEC/IEEE 42010 standard for architecture description) - yielding a continuous decision log
- Model tailoring (context-specific filtering), decision backlog management
- Rich text editing, model refactoring, reporting via Enterprise Architect integration
- Decision capturing with lightweight decision capturing templates such as Y-statements (as introduced in this IEEE Software/InfoQ article)
- Question, Option, Criteria (QOC) diagram support
- Sample guidance models compiling decisions that recur in cloud application design and workflow design
- Add-In to Sparx Enterprise Architect Version 10 (and higher)
- UML Profile and MDG Technology with state-of-the-art Architectural Knowledge Management (AKM) semantics, optimized for decision modeling with reuse
- Model tailoring and filtering capabilities based on Tagged Values (UML extension mechanism)
- Decision space analytics
- RESTful HTTP interface for tool integration
Vision, architecture and implementation of ADMentor are described in an IEEE/IFIP WICSA 2015 paper (author's copy) and supporting presentation (PDF). The users guide (use case walkthrough) can be downloaded here.
(Update 2016): ADMentor has been released as an open source project on GitHub.