When building dashboards it is often best practice to use the iCubes built by Digital.ai. These are pre-built datasets that incorporate metrics and attributes surrounding one functional area like "Incident" or blend multiple areas like "Change" and "Problem."
These cubes are rigorously tested and often provide the best performance and experience for building dashboards. These cubes are also upgrade-proof and maintained by Digital.ai.
It may be necessary to build a User-Created Custom Dataset or Report in order to fulfill your dashboard design requirements. A Custom Dataset is necessary for situations in which:
- A longer time filter is necessary than the 13-month one present in a standard iCube
- Blending cubes is no longer optimal for performance purposes
- Limiting Data / Left Outer Join support - This limits the data present in a selector, for example limiting the Assignment Groups available in a dropdown.
Please always work with your CSM to determine if a custom dataset is necessary for your scenario. Situations where it is not appropriate to create a custom dataset include:
- Attribute/Metric is not present in the iCube - work with CSM/Support to have this added instead of creating a new dataset
- Cannot Find appropriate cube - work with CSM/Support to understand the location of all available cubes
- Similar dataset/cube already exists - performance is optimized when datasets and cubes are reused as much as possible
Related Article: Guidelines for Dataset Creation in Digital.ai Intelligence