Splunk · Primly Community

Splunk coding interview / online assessment, format and difficulty breakdown

quietquit_quincy · 5 replies

Finished the Splunk online assessment two weeks ago for a mid-level backend SWE role. Here's what the coding interview / OA actually looked like, because I went in a bit blind.

They use HackerRank. The OA I got was 90 minutes, 2 problems.

Problem 1 was a string processing problem. Reading between the lines it was log-parsing adjacent, which fits the Splunk product context. Medium difficulty on the LC scale, maybe low-medium. I got it done in about 30 minutes including edge cases. They care about correctness on edge cases, not just the happy path.

Problem 2 was a graph / tree traversal problem. Medium difficulty, BFS/DFS territory. Nothing exotic. I used BFS, it passed. I didn't try to be clever with the approach, just clean readable code and handled the edge cases.

One thing worth noting: you don't hear back on the OA result through the platform. They just move you to next steps (or go quiet) via the recruiter. I found out I passed because the recruiter emailed 4 days later to schedule a phone screen.

A few tips from my prep: They don't seem to ask hard LeetCode problems in the OA. Focus on mediums, especially strings/arrays and graph traversal. Actually comment your code or at least name variables clearly. I've heard from people who did onsites that reviewers sometimes look at code quality. Time management: 45 mins per problem is plenty if you've practiced, but the 90 minute window goes fast if you get stuck on the first one.

For the onsite coding rounds I've heard it's similar difficulty, more conversational. But the OA is a pretty clean filter of "can this person write working code quickly."

5 replies

jp_newgrad

Thank you for posting this. I have my OA coming up next week and was worried they'd throw hards at me. Mediums feels way more manageable. Did the string problem involve parsing structured log lines or was it more of a generic string manip thing?

frontend_fran

Generic enough that you wouldn't need Splunk domain knowledge, but thematically felt like log-adjacent. Think parsing a formatted string and extracting structured data from it. Classic stuff.

bootcamp_bri

I did their OA last year and the format was similar. One difference: I got a slightly different graph problem but same BFS/DFS flavor. Consistent across time it sounds like.

sdr_sky

Did you use Python or something else? Wondering if the time limit is tight in Python for the test cases.

frontend_fran

Used Python. No issue with time limits on my end. The test cases weren't adversarially large.