A lot of discussions around this kicked off from https://github.com/apache/incubator-hudi/issues/869
Breaking down into phases, when we drop the hudi-spark-bundle*.jar onto the `jars` folder
a) Writing data via Hudi datasource should work
b) Spark datasource reads should work
c) a + Hive Sync should work
d) SparkSQL on Hive synced table works
Start with Spark 2.3 (current demo setup) and then proceed to 2.4 and iron out issues.