Databricks Interview Questions
& Process
Real candidates share what happened, how many rounds they had,
and how the experience turned out.
Based on 1,060 interview experiences · FREE TO READ
Which role are you interviewing for?
30 roles · 1,060 reportsCandidate interview experiences
First-hand accounts from people who interviewed at Databricks.
Databricks
AI Engineer
Mid Easy Negative experience No offer 5 rounds
Interview process
Phone screen Technical screen Onsite Onsite Background check
Interview formats
Technical System Design Coding Behavioral Other
Hiring manager chat, technical interview (knowledge about genai, system design, coding), behaviour round, final catchup, background check; bad experience as got rejected even after background check. Been told since this is their AI engineer - FDE role, they want to make sure 100% right as this will be foundation member and I don't have external consulting experience. Then why bother start at the beginning ?
Confirmed questions1 question
- - technical knowledge, system design, coding
Databricks
Software Engineer
Intern Average Positive experience No offer 2 rounds
Interview process
Technical screen Technical screen
Interview formats
Coding Technical
there were 2 stages: coding and algorithms. in the coding stage i had to write methods and tests for a LazyArray, and in the algorithms stage there was a problem where i had to use set (i was writing in c++) but the interviewer asked me how else could i implement the code with all possble data structres, array, heap etc and analyse runtime of these alternative solutions
Confirmed questions1 question
- implement two functions for a lazy array
Databricks
Software Engineer
Entry Average Neutral experience No offer 3 rounds
Interview process
Phone screen Onsite Onsite
Interview formats
Technical Coding System Design Behavioral
Applied as a new grad with a referral, had one phone interview, then onsite with 2 systems interviews and then one behavioral round. I applied as a new graduate with a referral. I had one phone interview, followed by an onsite interview that had two systems interviews and then one behavioral round.
Confirmed questions3 questions
- House robber for phone, then IP to CIDR, and designed a cache and multi-threaded system
- House robber problem for the phone interview, then IP to CIDR, plus designing a cache and multi-threaded system
- Designing a cache and multi-threaded system after doing House robber for phone and IP to CIDR
Databricks
Account Executive
Senior Difficult Negative experience No offer 4 rounds
Interview process
Phone screen Technical screen Group
Interview formats
Behavioral Technical Presentation Group Other
The hiring process was extremely long, difficult, and overly demanding, requiring unreasonable amounts of research and data preparation that really seemed like the actual scope of the role. It was hard to interpret virtually, with several tense, recorded one-on-one interviews, and these were further disrupted by senior management not showing up at scheduled times. The overall environment felt perfunctory, making me think they already had an internal candidate and were just proceeding for elimination or backup. The final interview was held virtually in a large group setting—a reasonable format for a sales role—but there was no mention of any in-person component. This included demands for very deep dives into technology topics that usually need an engineering resource to provide credible data, putting the sales candidate in the awkward position of maybe having to speculate or "make things up." I also noticed signs of serious bias: the team lacked racial diversity (everyone appeared to be white/Caucasian) and they made negative remarks in 1:1 meetings that hinted at ageism and racial factors, specifically aimed at candidates from certain technology companies or backgrounds.
Confirmed questions1 question
- Please walk through Databricks' value proposition, outline how its architecture differs from a competitor like Snowflake, and explain the main competitive advantages that help Databricks win in the market.
Databricks Interview Questions
Quoted word for word from Databricks interview reports.
“What are the assumptions behind linear regression and what happens when those assumptions are broken?”
Read reports →“Q2: when forecasting average revenue, would a random forest tend to overestimate or underestimate revenue compared to an ARIMA model?”
Read reports →“How would you go about troubleshooting a Mac that can’t connect to WiFi?”
Read reports →“You are given a tree (all edges known) and a set of target nodes for a person starting from the root node (list of vertices given). What is the minimum distance required to visit all target nodes and then come back to the root? Follow-up: Now there are 2 friends, each with their own independent target nodes. What is the minimum total distance travelled by both friends (distance of friend 1 + distance of friend 2) to visit their targets, given that at any point the maximum distance between them can be the length of 1 edge (if one person is at node x, the other can be at the same node, its parent, or one of its children)? Additional follow-up: Same as the follow-up, but now the maximum allowed distance between them is at most k (integer k given).”
Read reports →“What is considered best practice for data storage for ML (lake or warehouse)?”
Read report →“How would you remove PII from a data lake when you get requests to delete it?”
Read report →“Can you explain what a data warehouse is and how it differs from a regular database?”
Read report →“Create a trigger on the Account object that updates fields based on its related list, and for a scenario with 3 levels of hierarchy, solve it using a queue.”
Read report →“How would you design a notification service that can handle both real-time and batch deliveries?”
Read report →Formats, difficulty and experience
Across all 1,060 Databricks interview reports.