Of course, no one can just start building use cases on a technology platform that does not exist yet. Things have to be architected and designed. Architecture, design and having a plan for those is important. And there are many types of architecture involved. But generally we can split it up into two rather independent streams of work as mentioned above: the platform and the use cases it supports. The platform is primarily a technical concern whereas the design of the use cases is primarily about understanding the business.
Once the business case has been approved, as it comes out of the Chart phase an initial team can be created, the scope is clearly defined, and we know what the first use case is to start with. First thing to do then is to organize design sessions with that team and go through all the various types of architecture and devise a project plan of milestones and clearly defined deliverables. Avoid doing it the waterfall-way, do it Agile but Agile does not preclude you from first having a good plan.
- A well-informed team, prepared to successfully implement your use cases.
- More detailed use case tree, including datasets, data sources, ontologies, personas, business outcomes and user stories, all delivered within the context of the UCT, as a model.
- Then the translation of that to:
- Ontology Architecture
- Technical Architecture
- EKG/Platform, identification and functional description and requirements of all the various platform services needed.
- A design and architecture for the ”DataOps Environment” which connects the upstream data providers and downstream data consumers to the EKG/Platform.
- A design and architecture for the user experience (either bespoke or generic model-driven UI).
- A clear sequence of deliverable artifacts as input to the project plan.
- A realistic roadmap that can be explained satisfactorily to stakeholders