Yelp ML Engineer interview typically runs 4 rounds: offline coding challenge, recruiter chat, live coding, and a 4-hour final panel. The process usually takes a few weeks and is notably structured and lengthy.
$90K
Avg. Base Comp
$155K
Avg. Total Comp
4
Typical Rounds
2-4 weeks
Process Length
We’ve seen Yelp lean toward candidates who can make a simple problem feel crisp, not flashy. In the experience we reviewed, the standout technical prompt was Jaccard similarity between sentences — straightforward on paper, but revealing in how cleanly the candidate decomposed it under time pressure. That’s a recurring signal for Yelp: structured reasoning and clear communication matter more than exotic ML depth in the moment.
A second pattern is that Yelp seems to care a lot about how you operate in a mixed-format interview day. Our candidate described the recruiter as warm and organized, and one manager conversation as genuinely kind and attentive, which suggests the company does value professionalism and collaboration. But the same report also noted a system design conversation that felt dismissive and overly critical. That contrast tells us candidates are being evaluated not just on correctness, but on whether they can stay composed when the interaction gets less collaborative.
We also see a company that compresses a lot into a single stretch, so endurance becomes part of the signal. The final panel combined behavioral, design, and coding, and the candidate called it exhausting. In our view, Yelp is looking for people who can keep their thinking sharp across different modes without losing clarity. The non-obvious make-or-break factor here is not just solving the problem, but doing it in a way that still reads as practical, calm, and easy to work with.
Synthetized from 1 candidates reports by our editorial team.
Had an interview recently?
Share your experience. Unlock the full guide.
Real interview reports from people who went through the Yelp process.
The part that stood out most to me was the live coding question on Jaccard similarity between sentences. It was a straightforward idea, but I had to think carefully about how to break the problem down cleanly under time pressure. That ended up being a good preview of the whole process at Yelp: not overly exotic, but definitely structured to see how you reason through a problem and communicate while doing it.
The process started with an offline coding challenge, then moved to a recruiter chat. The recruiter was warm, engaging, and professional, and they laid out the next steps clearly, which made the process feel organized early on. After that I had a live coding session with one of Yelp’s engineers. It was interactive and gave me a chance to talk through my approach in real time. The final stage was a 4-hour back-to-back panel interview, which was pretty exhausting. It included two behavioral interviews, one system design interview, and one live coding interview. One of the behavioral conversations was with a manager who was genuinely kind and attentive, and that part of the day felt very positive. The system design round, on the other hand, felt less collaborative than I expected, and the interviewer came across as pretty dismissive and overly critical, which made that section harder than it needed to be.
Overall, the technical difficulty was manageable, but the length of the final panel made it tough to stay sharp the whole time. I also asked for feedback afterward and never received any, which was disappointing. I ended up not getting an offer. If you’re preparing for this process, I’d make sure you can implement Jaccard similarity quickly and cleanly, and be ready for a long final panel that mixes behavioral, system design, and coding in one stretch.
Prep tip from this candidate
Be ready to implement Jaccard similarity between sentences from scratch, including the tokenization and set logic, since that was the concrete coding question. Also prepare for a long 4-hour final panel that combines behavioral, system design, and another live coding round back-to-back.
Share your own interview experience to unlock all reports, or subscribe for full access.
Sourced from candidate reports and verified by our team.
Topics based on recent interview experiences.
Featured question at Yelp
Determine whether there exists a permutation of an input string that is a palindrome.
| Question | |
|---|---|
| The Brackets Problem | |
| Valid Anagram | |
| Flatten N-Dimensional Array to 1D Array | |
| Groups of Anagrams | |
| Sort Strings | |
| Reservoir Sampling Stream | |
| Target Indices | |
| Find Mismatched Words | |
| String Palindromes | |
| Common Prefix | |
| Evaluate News | |
| LRU Cache 1 | |
| Intelligent Restaurant Review | |
| Sum Numbers As Strings | |
| Length Of Longest Palindrome | |
| Merge Sorted Lists | |
| Weighted Keys | |
| String Shift | |
| Bagging vs Boosting | |
| First to Six | |
| Job Recommendation | |
| 500 Cards | |
| P-value to a Layman | |
| Hurdles In Data Projects | |
| Maximum Profit | |
| Compute Deviation | |
| Raining in Seattle | |
| Nearest Common Ancestor | |
| Get Top N Frequent Words |
Synthesized from candidate reports. Individual experiences may vary.
The process begins with an offline coding challenge. It appears to be a structured screening step focused on practical problem-solving before any live interviews.
Next is a recruiter conversation where the recruiter explains the process, answers questions, and outlines the upcoming interview stages. The experience was described as warm, professional, and organized.
Candidates then complete an interactive live coding session with an engineer. In this interview, the question centered on implementing Jaccard similarity between sentences, with an emphasis on clean decomposition and communication under time pressure.
The final stage is a back-to-back panel interview lasting about four hours. It includes two behavioral interviews, one system design interview, and one live coding interview, making it a long and demanding final round.