Salesforce USA SMTS Offer
Anonymous User
811
Interview

I interviewed for Senior MTS role

  • Hiring Manager Screen (45 mins)

    • Questions related to my work experience, my role + impact
    • Behavioral questions for almost 45 mins. We went over 1 hr for the interview duration.

  • Online Assessment (60 mins)

    • 4 standard leetcode problems. The last one boiled down to finding the longest increasing subsequence which I could not code in O(nlogn) due to which I could not pass 7-8 test cases.

  • Onsite

    • Coding (60 mins)

      • 1-2 Behavioral questions
      • There was a standard linkedlist problem regarding reverse sublist from 'fromIndex' to 'toIndex'. Some simple followups + optimization discussions.
      • Additional discussions on some system design aspects mostly around the cloud service providers I have worked on (AWS). Discussions on DynamoDb, Lambda and SQS.

    • System Design (60 mins)

      • 1-2 Behavioral questions
      • This started with a normal question answer type session. e.g. What is a REST Api, what is HTTPS, what is JWT etc etc. There were a lot of questions asked to understand the depth of knowledge.
      • We only had 20 mins left for system design question. The question was to design a service similar to TicketMaster. The interviewer grilled me after every sentence I was speaking. We had a lot of in-depth conversations on how atomicity is ensured in an ACID compliant db (literally going at the thread level). Why does Elastic Search speeds up search queries, what is inverted index, what are some problems with Elastic Search etc. We actually went 25 mins overtime. Not a very professional thing to do and horrible candidate experience.

    • Coding / Design (60 mins)

      • 2 Behavioral questions
      • This was again a question answer type of session. Started with what is the difference between Arrays and Lists, What is caching, how to implement a hashmap, what is caching, what happens if you cannot fit the cache in memory (cache evication), how is LRU implemented, how is LFU implemented. Then some design discussions on a problem related to Stock Broker app. What is the best way to show historical prices of a stock to the client? I mentioned time-series db. Pros/Cons of it, why it speeds up time-related queries. How to handle stock ticker changes on the app etc etc. I mentioned pub/sub here instead of regular polling. What is a Pub/Sub, how does Redis pub/sub work? These questions were more of knowledge testing and were mostly checking breadth and high level understanding. Another round of discussion on KAFKA, its internals, message brokers in general etc etc. I did not write a single line of code in this round which was super weird. In my opinion, it felt borderline insulting to ask questions related to arrays and lists, hashmaps etc. These are ok to gauge a university grad, not a senior/staff candidate. It was an interview with no script in my opinion.

    • Behavioral (45 mins)

      • Pick a project of your choice, explain design choices, explain your role impact, what challenges you faced etc etc. Bunch of standard behavioral questions, calculated risk you took? Have you ever failed in your projects? Example of org wide / team wide impact, any processes you changed in your team ..

    • Coding (60 mins)

      • 1-2 behavioral questions
      • Hackerrank problem, big prompt but nothing too difficult. Took me some time to understand the problem and sample inputs and expected output. They ran with all the added tests and everything passed.

Overall, the most unstructured interview rounds and some really meaningless questions which should not be asked to a senior/staff person.

Recruiter reached out the next day saying the team would like to extend an offer.

Preparation Post: https://leetcode.com/discuss/post/7368458/coupang-intuit-salesforce-oracle-c3ai-se-n0gq/

Comments (2)