Some Useful Tags:
Pre-Placement Preparation, PPT’s, Resume, Tests, Interview Preparation, Placement Experience, Goldman Sachs, Tower Research, IBM, Uber, JP Morgan, Microsoft
Profile at a Glance:
1) CPI : 8.83
2) Minor in Stats, Honors in CS
3) 4 internships (2 Tech, 2 Data Science) + 1 RnD project
4) Lots of Projects
Pre-Placement Preparation:
1) PPTs + Seniors Talk :
You will have to attend a fixed number of PPTs due to the placement policy. Do attend PPTs of companies that you are interested. There will be a lot of information and most of the companies will talk about their business etc. but the important information to take away is what the company does and what would be your role in the company. Apart from this, you should get feedback from people who are already present in the company, either through sites like Glassdoor or by contacting seniors who are placed there. You can get this info (seniors) from the previous years placement blogs.
2) Resume :
Your resume is a very important factor both for shortlisting for the interviews and during the interviews. From my experience, CS Tech companies are mostly interested in your internships as well as your projects. Most of my interviewers didn’t even go to the second page of PORs & Extra-currics, so try to make your resume in such a way that most of the things you want to highlight are mentioned in the first page itself. Have a look at some of your senior’s resume once before making yours. Finally complete your resume well before time and send it to as many friends/seniors as possible and get feedback and make improvements.
3) Tests :
This would be a major factor for getting shortlisted for the interviews. Most of the Tech companies tests are purely coding based having 2-3 questions in typically 1-2 hours along with some MCQs on basics of Computer Science. You should practice coding questions from sites like Hackerrank and Codeforces since most of the questions are similar in nature to them and those questions will help you get your basics clear. The more the questions you do, the better since there is a high chance that a similar question in principle might come in the exam. Apart from this, for finance companies you should practice your probability and statistics skills and do a lot of puzzles. Remember though, that the test along with the resume are used for shortlisting the candidates for the interviews by most of the companies. So don’t lose heart if your test went bad.
4) Interview Preparation :
First and foremost, go through your resume well, each and every point. You never know what might catch the interviewer’s eye and you shouldn’t be left thinking about it in the interview. For projects, prepare justifications for design choices and also revise the mathematical details of the methods used in your project. Apart from this prepare for the basic HR questions. Lastly prepare some questions you would like to ask the interviewer. Also think about questions like why this particular role?, Why this company?. A major portion of my interviews was on these questions in finance companies like GS and JP Morgan. Finance companies ask a lot of puzzles and probability and stats questions in the interviews, prepare well for those. And for CS Tech companies, coding on paper is common. Go through resources online for pointers regarding preparing for the above. One important thing is to always think aloud, even if you are stuck at some point. The interviewers are mostly interested in the way you think rather than whether or not you get to the correct answer.
Placement Experience
I was shortlisted for 8 companies on Day 1, 4 in each slot. The placement cell assigned ½ hour slots for each of the companies to conduct their first rounds of interviews. But these slots were hardly followed beyond the first hour. So it is essential that you have friends on that day who will coordinate the timings of your interview slots with the company coordinators and ensure that your time is not wasted and you are able to attend as many interviews as possible. Also be prepared to make quick decisions on where to go, which may also involve choosing between 2 companies due to time constraints. A day or two before, run through different scenarios and make up your mind about what to choose. For e.g. in my case I had to forgo going for the Google interview altogether in order to give the remaining 3 interviews well in the D1S1 and same with Apple in D1S2. There are typically 2-3 rounds of interviews for most of the CS Tech companies and also for some of the finance companies, and it gets really exhausting till the end of the day. In my case, my first interview for Tower Research (eventually this is where I was placed 🙂) took 2hrs and then I had to rush through everything else. As a general rule if a company is interested in a candidate they try to increase the length of the interview.
Tower Research
1) Resume based shortlisting for the interviews.
2) Interview :
First Round : Intro, Questions on my projects, Networks based questions, O.S. based questions, coding questions. Asked to code up on pen-paper. (50 mins)
Second Round : Similar as the first round in the initial discussions. More focus on coding question (graph based), asked to code it up on pen-paper, asked about my PORs. (50 mins)
Uber
1) Coding Test : 3 Coding Questions (1 1/2 hrs). 2/3 for getting shortlisted.
2) Interview :
a) First Round (Tech) : Asked about how to model the tetris problem and code it up on paper. (40mins)
b) Second Round (Tech) : Asked coding questions (3), Puzzles, and then asked to code up the solution for one of the coding questions. (40mins)
c) Third Round (HR – based) : Intro, Motivation, Questions on internships testing your involvement (Basic Questions like how big the data was, some of the challenges that you faced while coding it up, the systems used to handle the huge amount of data, etc.), Biggest Failure and how you have overcome that failure, Strengths. (40mins)
Microsoft (Software Developer and Data Science)
1) Test :
a) SW : Coding Test; this included 3 easy questions. The challenge here was that the environment was their own cocubes whose interface is really difficult to use and debug. So essentially they were testing how well can you code on white paper i.e. without the debugging.
The questions were testing skills with pointers, trees, etc. (1hr 15mins)
b) DS : This test included questions in details about various Machine Learning models. Some of the questions were factual but most of them were application oriented. for e.g. effects of various hyperparameters, which loss function to use, etc. (1 hr)
2) Group Fly :
a) SW : This stage included 2 questions to solve on pen paper, and then discuss with the guides there to optimize the solution if possible. One of the questions was based on binary trees and the other was on RLE. (1 ½ hrs).
b) DS : 2 Questions – The first was to explain the mathematical formulation, learning algorithm and prediction algorithm of any of the following : (Neural Networks, HMMs, MRFs, LSTMs, etc). The second question was on practical applications of ML. 4 ML tasks were provided, we had to formulate any one of the tasks as an ML problem, ask for the required data and propose a model which was able to solve the problem and discuss the nuances of the task with the guides present. (2 hrs)
3) Interview :
a) SW :
1) First Round (Tech): 2 coding question on binary trees. Then a simple tweak to the question. Then he asked me to code it up on pen-paper in C++ (1/2 hr).
2) Second Round (Tech) : Discussion about my projects, motivation, talk about why not go for MS? Then a coding question again on binary trees. Asked me to code it up on pen-paper in C++. (40 mins)
3) Third Round (Redmond) : Didn’t get shortlisted for this round.
Goldman Sachs
1) Test : The test was mostly testing your JEE Math skills and some concepts of statistics and probabilities and a few simple questions on coding algorithms, along with one coding question which involved solving a puzzle itself (not much weightage given to this).
2) Interview : There were 3 separate teams that had come for the interviews and each of them had a similar but independent process.
1) First Round : Intro, Questions about motivation (Really Grilled me here), why finance, why GS?, etc.
2) Second Round : Intro, Motivation, Stats based questions on variance and derivations for simple time series data.
IBM (Data Science Profile)
1) Test : 3 rounds of test, one on Logical Reasoning, Data Interpretation and one on English. (2 hrs total)
2) Interview : One HR based interview with questions focussed on ML projects.
JP Morgan (Quantitative Researcher)
1) Test :
a) MCQs – 3 sections : Probability and stats based questions, Maths based Questions, and CS Coding based questions (1hr 15mins)
b) Coding – 2 coding questions on binary trees(½ hr)
2) Interview :
a) First Round : ML questions, mathematical details on concepts used in projects.
b) Second Round : CS theory based questions on Compilers and OOPs in C++.
c) Third Round : Questions on OOPs in C++.
Be-Prepared 