Hi Folks,

I have around 10 years of experience with 5 in DS/ML domain. Currenly, I am working in a leading consulting firm in DS/ML.

Sharing the interview experience with the multiple companies I have recently interviewed with -

• Cloudera - Staff ML | REMOTE

  • Round - 1 : DSA | Question was to find the top k jobs which are taking more time.

    • Interviewer provided a class for JobScheduler with members like job_id, start_time, end_time etc.
    • I tried to comeup with solution with calculating the duration of each job and then try to find the top time consuming job which are taking more time. --> The solution was not good and I messed up when asked about TC for the solution.
    • Provided one more solution to avoid storing the duration seperately in a dictionary and instead based on job which in considered, check if it is one of the K top most time consuming job.

  • Round - 2 : ML System Design | Design an anomaly detection system where we need to detect anomaly around cpu_time, ram consumption of the job etc.

    • IMO, it went fine, I was able to put out my thought process on how I will design the solution including monitoring etc.

  • VERDICT : REJECTED | It was good opportunity, felt bad to mess it up.

• Bain- Advanced Analytics Group - Lead ML | Hybrid, GURGAON

  • Round - 1 : DSA | The round was taken on codility. I dont recall the exact questions but they are Leetcode- Easy. For the first question, we just discussed the solution and for second, coded it and all test cases passed as well.
  • VERDICT : REJECTED | Not sure what went wrong! It was also a good opportunity with AAG.

• Nagarro : Associate Principal Engineer - ML | REMOTE

  • Round -1 | OA around on quant of around 20 minutes | You can refer any youtube video to get an idea.
  • Round -2 | It was mostly technical discussion and I was able to address most of the questions, but had no experience in Computer Vision domain.
  • VERDICT - REJECTED | Recruiter told, I am not having all skills needed for the job but whatever skills I have, they are good. Also, there was a concern around Expected CTC too. Later, I got a rejection mail.

• General Mills : Lead ML | REMOTE

  • Round -1 | OA around python, DS, MLOps, SQL | It was difficult.
  • Round -2 | It was technical discussion with a Third Party Recruiting Firm member. Technical discussion went fine and I felt the interviewer was taking too much time to ask questions or was not sure what too ask. Finally, he gave me two questions which was not able to solve with optimal solutions-
    • SQL : Write a SQL query to get the 4th highest salary from Salary table without using LIMIT, TOP.
    • DSA : Provided a list, split it into two list with same sum --> I was not able to handle all cases.
  • VERDICT - REJECTED

• Thoughtworks | Senior ML - Hybrid, Gurgaon

  • Round -1 | HR shared a github repository and asked to go through the code base. During interview, the panel member asked to build a user story around train the model and register it in Model Registry (MLFLOW) and then use it for inference via API written on FastAPI and that API needs to be used in Streamlit Application which takes in user input and returns the response.

    • The round went awesome and interviewer was highly impressed with the details I had put out while solving the provided problems.

  • Round -2 | It was a Technical Discussion (90 minutes but ended in 60 minutes) with AI Lead. Interviewer was not fully convinced with the answers I had provided. I went with a mindset to solve engineering problems but I was taken-back by too much focus on inner working on python dictionaries and ML Algorithm --> It was good experience, honestly because once you have around 10 YOE, it becomes difficult to keep in mind all the things. Right now, I feel like I need to know DSA, LLD, HLD, ML, Gen AI, MLOPs etc and some things are bound to spill out. I am working on fixing the loose ends in all these areas.

  • VERDICT - REJECTED

• TCS | AI Lead - Hybrid, Bangalore

  • Round -1 | Technical discussion for 60 minutes.
    • The questions were around classical Machine Learning algorithms, evaluation and NLP including below coding questions -
      • Valid Parenthesis | All test cases provided by interviewer were passed. Interviewer was keen on making the solution fail :-D.
      • NLP based ML Problem | I was asked to build ML pipeline for NLP classification problem from the dataset provided by sklearn datasets | Since we were short on time, I was able to code preprocessing part and explained the next steps, which I would do to build the Train Pipeline.
  • VERDICT - REJECTED
    • Not sure what went wrong, since I was able to precisely answer all the questions. On NLP, I was bit rusty, but I was able to provide logical solution to the questions. Interviewer seemed convinced too. Tough Luck!

• Airamatrix | ML Engineer | Mumbai, Hybrid

  • Round - 1 | 1 hour discussion with two members
    • On MLOps side, I was not able to answer the questions by the members. It looked like the team was looking for more hands-on on Kubernetes/Docker etc.
  • VERDICT - REJECTED

• Persistent Systems - AI Architect - Gen AI, Bangalore, Hybrid

  • Round -1 | 30 minutes discussion with HM.
    • The discussion went around scenario based questions, Gen AI, DS/ML.
  • VERDICT - ON-HOLD
    • HR told the position is on hold for now.

Cheers! All the best for your preparation!