Debdeep Paul Chaudhuri: Placed at United Airlines

Debdeep Paul Chaudhuri(21111413), MSR Y21, CSE, studying in IIT Kanpur will be joining Company - United Airlines Profile - Associate Data Scientist. Here are some brief insights.

Shortlisted Profiles & Companies:

1. SRIB, Senior Developer 2) Axtria, Analyst

Selection Process of the Company you are selected in 1) Resume Shortlist, 2) GD, 3)Test

For Online Assessment round(90 mins), there were 3 sections 1) A MCQ based section comprising questions on fundamental ML algorithms, stats and probability, medium aptitude. 2) 1 or 2 SQL query questions. 3) A leetcode medium dsa question.

Insights on the interview process

2 rounds (both technical) 1st round 30 min 2nd round 1 hr.

Sample Interview Questions?

1st Round(30 mins) focused on basic python programming skills and SQL query logic building skills, some SQL theoretical questions,followed by a few ML questions and it lasted for 30 minutes. Python program was as basic as coding fibonacci series upto nth term. A medium SQL query question, where they mainly focused on the logic building ability, as I didn’t remember the syntax but could partially solve it with the hints that they were dropping, I was considered. Some theoretical SQL questions like GROUP BY vs HAVING clause, Aggregate functions in SQL and so on. ML questions were based on Decision tree, bagging vs boosting(Very imp. Concepts from interview point of view). 2nd Round(60 mins) was divided into 2 segments 1st was an ML breadth and depth segment(30 mins), 2nd was a case study segment(30 mins). In the 1st segment the interviewer asked all sorts of ML questions like Ensemble learning(again v. important), generative modeling vs discriminative modeling, evaluation metrics like precision and recall and what would be used in a spam filter,regularization techniques etc. and DL questions on gradient descent, backprop, objective functions that are used in LLM(This was asked as I had been working on LLM for my thesis, plus it is the hot topic of today’s time), role of filters and pooling layers in CNN and so on. The 2nd segment was a case study on Zomato food delivery. This was more of a discussion session, where they didn’t want the exact solution but wanted to figure out what was my thinking process to approach the solution. The initial discussion revolved around data pre-processing/engineering, like what features to be kept and discarded. And after about 20 mins they emphasized on the model building approach.

Preparation Resources

1. DSA from Striver sheet 2) Aptitude from Prepinsta, gfg 3) SQL from Leetcode 4) ML and Data Science (CS771, CampusX, blogs in towards data science, analytics vidhya, machine learning mastery are notable few, some design/case study questions can also be prepared)

Advice for students (do's and don'ts)

1. Prepare ML and DL basics really well, as for fresher hire they drill you on the basics. 2. SQL is a must if you’re aiming for DS roles ( I regret not preparing for it enough!!). 3. Think of bridging theory and practice, again this will help in case study rounds. 4. Some good projects demonstrating your knowhow and expertise always helps. 5. Do keep a backup plan in your head(even if it’s hypothetical), which will relieve a lot of stress during interviews. (A small mental hack)) 1. Don’t solely prepare for Data Science roles even if you’re eyeing for one, rather keep the preparation general(by that I mean prepare for 2 other roles, which will give you an opportunity to face more interviews), and you can always switch/hunt for your preferred role later, while having an offer in your hand. 2. Don’t let disappointments and despair get the better of you. Even if you see your mates getting placed while you’re struggling, try holding onto the idea that “you’ll sail through somehow”. YOUR TURN WILL COME, AND SOME OF THOSE MATES WILL BE YOUR GREATEST CHEERLEADERS!!