One of the focuses of AgileBI should be removing latency from the entire delivery process.
We do this by identifying patterns we can use to reduce the effort and risk in doing repeatable tasks. And then when we have these patterns identified, we should automate as much as possible to again reduce the effort and risk in executing these tasks.
For the foreseeable future there will always be a human required to under take some of these tasks. These should still be undertaken using patterns, but not automated ones.
== latency in the agile process
[[bringing the concept of lean into AgileBI]]
[[in fact mature teams merge scrum and Kanban into a single lean process]]
== agile data engineering
[[auto heal or fail in art to remove the latency of breaks and fixes]]
[[test driven deployment, remove latency in testing]]
[[load all data first into time variant reduces latency of adding new data to business layer]]
[[define and apply test at the same time you apply the business rule]]
{{there is a push in the data vault automation world at the moment to build software that automatically builds the raw vault structure based on analysing the data structures in the source data. Using foreign keys to self determine the hubs etc.
[[good wizard to show what might work]]
}}