What is the desired functionality that you are missing?

As a User I need a way to update process variables outside of the usual import mechanism. 

Which problem are you going to solve with this functionality?

There are different use-cases where this can solve problems:

a) Only business-keys that are relevant in a particular Camunda process are stored inside the Camunda Engine. All other information is getting stored in external systems. During the variable-import via org.camunda.optimize.plugin.importing.variable.VariableImportAdapter  all needed business data can be imported. Unfortunately, this is only done once, as optimize imports the variable after it has first been created (when the process instance is started). After the initial import, Optimize only fetches Variables that have been updated since the last import. In this case, there is a new entry in ACT_HI_DETAIL. The business-keys however is never getting updated, so it never updates the initially imported business data in Optimize. 

b) In some environments not all variables that are relevant in a particular Camunda process are stored inside the Camunda Engine but rather in external systems. In these cases variable data is not visible in Optimize, but rather a reference is stored in Optimize. Sometimes this data in external systems is being updated without the knowledge of the Camunda Engine, but still it is required that the business data is being updated in Optimize too. 

Context

To enrich the imported data from the Camunda Engine History with this data it is thus desired that Optimize offers a functionality that allows third-party systems to inject this data and enrich the process instance data in Optimize with additional variable information.

AT:

• a scalable way to inject this data is designed and implemented (as a user I probably pass in a variable (or multiple variables) and a process instance id
• the approach is documented in the technical guide