I applied to Databricks through their website and later(after about one month) received a referral from a friend. A day after the referral, a recruiter reached out to me and scheduled a 15-minute call. During the call, the recruiter asked about my experiences listed on my resume, my interest in Databricks, and the specific position I was applying for. It's crucial to thoroughly understand the role you’re applying for, especially after receiving a referral. For example, expressing interest in AI and ML while applying for an infrastructure position may not align well with the recruiter’s expectations.

After this, I was scheduled for a 1-hour phone screen, where I was asked a typical Databricks medium-level question. The interview went smoothly, and the next day, I received a link to schedule a virtual onsite interview.

The onsite interview consisted of four parts:

1. Algorithm Interview: This was similar to the phone screen(LC medium) but involved two interviewers.
2. Cross-Functional Interview: A typical behavioral interview where I discussed my experiences.
3. System Design Interview: This involved designing a simple system, discussing scalability, trade-offs, and other considerations.
4. Coding Interview: The goal here was to design a class and implement it while working with the interviewer. I struggled with this part, mainly because I hadn’t found many references or resources to prepare effectively.

A few days later, I received a rejection. The first three interviews are managable, but the last one was challenging due to the limited reference material available for preparation. But my advice for the coding round is, know your language very well - classes, objects, funciton chaining implementation(your_object.foo().bar()).

Let me know if you have a question, but couldn't share specifics because of the NDA.

UPDATE : Looking back on the 4th round, The questions looks like Amazon's OOD/LLD questions.