OLAP Cube is not dead, as long as you use some form of: 1. GROUP BY multiple fields (your dimensions), or 2. partition/shuffle/reduceByKey and materializing/

OLAP Cube is not dead, as long as you use some form of: 1. GROUP BY multiple fields (your dimensions), or 2. partition/shuffle/reduceByKey and materializing/caching the result of this query for later reuse, you are rolling your own OLAP Cube with whatever data processing ecosystem you have on hands. The technology itself has clear use cases and incredible business value on a daily basis, only implementation details differ depending on surroinding ecosystem