So thanks to the various comments on my posts - they were helpful.
What I have now is:
druid ingestion from kafka
queries whose results provide evidence druid is receiving the data from kafka
So all good.
My counterpart is building some js web front ends to visualize the analytics we are doing on our druid data.
What we need are methods to access say the most recent interval x of data from druid, from java script ( we are using react and node.js).
Then we need to stream updates as they occur and roll the interval x along as updates occur.
I figure this is something done a million times… does any one have any suggestions that might point us in the right direction?