I applied in-person. The process took 1 week. I interviewed at SnappyData in Aug 2018
Interview
Hands-on Coding
Personal Interview (3-4) rounds
Focus on concurrency, distributed systems problems, data structures that you have used not what you know theoritically, spark internals, transaction semantics (mvcc or acid).
You should really understand usage of Kubernetes if you say so (viz. resource allocation, network usage monitoring) linux skills should be average (not completely novice).
Interview questions [1]
Question 1
Implement ConcurrentHashMap without using standard java collection