Exciting company and space. Interviews were difficult, and the process was long. The recruiter, though mostly impersonal, kept me updated throughout the process.

The phone interview was a coding round with a medium/hard difficulty graph question. A similar (not the same) question is tagged for Airbnb on LeetCode, so I had seen the pattern before.

Onsite included a concurrency round where reading Chapter 3 of DDIA came in handy. I also found some Dropbox concurrency interviews online, which were sufficient preparation for this round. I had to spend several days preparing for this round, as it is unique to Databricks.

For the system design interview, key things to know were message queues, concurrency, and idempotency of transactions.

Onsite coding rounds were quite difficult. One was the same as frequently tagged on LeetCode for Databricks. Another relied heavily on knowing how Spark works conceptually, which sadly I didn't.

Some interviewers were very nice; some interrupted way too much and didn't let me think. Mixed experience.

BFS/DFS IP / CIDR