Goldman Sachs Interview Experience | Set 22 (On-Campus for Analyst profile)
I would like to share my GS interview experience with the prospect of helping upcoming batches.
Round 1 –
Resume based shortlisting – Basically add your projects related to Machine Learning and coursework related to Math. GS is one of the companies which reviews resume carefully (At least I think so :P), reason being me selected even after a low CPI. Around 250 to 300 students were shortlisted from resume for second round.
Hey geek! It's time to become a success story instead of reading them. Check out our most renowned DSA Self Paced Course, now at a student-friendly price and become industry ready. And if you are looking for a more complete interview preparation resource, check out Complete Interview Preparation Course that will prepare you for the SDE role of your dreams!
Feeling prepared enough for your interview? Test your skills with our Test Series that will help you prepare for top companies like Amazon, Microsoft, TCS, Wipro, Google and many more!
Round 2 –
Test on HackerRank. It included three sections with 45 min for each section. Coding questions (+20, -0) were to be solved in hosted compiler, some were integer type questions to be filled in the blanks (+10, -0, upto 2 decimal places) and some were MCQ (+10, -3).
Programming (Networking, Coding) – DSA revision would go a long way here, believe me, as there were direct questions on complexity of well-known algorithms and postfix, prefix expressions (Stack). 2 coding questions were asked. Be careful, optimized code is the way to go here as your solution may result in TLE for most examples.
Quant – Questions on basic aptitude like cattle grazing a circular field tied at a pole with rope and dice rolling probability were asked. Also some questions on expectation value, covariance, correlation and some matrices, determinant problems. I would recommend revising Prob Stats and Linear Algebra course if time permits.
Data Science – ML theory will help a lot. Questions were quite conceptual as they demanded theoretical knowledge like what to do to compensate for missing data. Some of them had 4 statements and correct ones have to be selected.
You have to do well in any two of the three sections,but an overall high score will increase your chances of getting selected easily by smooth interviewing. Also interviewing preference is set on the score in test. Here around 60 students were selected for next round, but beware as most of them are selected just in case. Because some of my batch mates couldn’t get another round of interview after their first mistake. When it comes to tests, higher you score, the better to leave a little load for last rounds (Some of us aren’t good at conversations after all).
Round 3 –
Here comes the loads of interviews at 6:00 am. Around 5 to 6 interviews if you get lucky. Although lesser number doesn’t mean you are rejected. They try to find the best fit department for you. Be punctual guys. Believe me they are there on time. And get a good night’s sleep. I tried to sit in as many companies as possible. I had a test up till 2:00 am that night. I was sleepy head all the time (they were considerate and sympathized for me, worked in a positive way for telling hard working attitude) as I couldn’t get proper sleep and had to wake up at 5:35 am (1st December chills of Kanpur) to check interview schedule and rush to venue meanwhile tying knots and combing hair, putting some deodorant and assuring my mom that I have woken up and trying not to fake it at all.
1st interviewer asked me a question on convex hull. Basically write a code to connect points in a way that they generate least area closed figure and encompasses all the given points. At first I couldn’t answer it. 6:00 am and you are asking me convex optimization, you must be kidding, right!!! (although it is a standard problem which I came to know about a little later in canteen with my friends who told me that it was so easy and I had to tell that hell yeah! I passed it with flying colors.)
Then he reverted back to a simpler problem. Good old problem playing number game with your friend to select any of the number between 1 to 3. Whoever reaches 20 first, wins. You have to tell the strategy to win the game. Basically, you start with your position at 20 and recursively think the problem to ensure that you end up at 16 (problem specific) before so that your friend ends up at any number between 17 to 19 and so on solve for 16.
2nd interviewer asked me a question on calculating variance, about gaussian distribution and something about my internship.
They won’t let me go for other companies in between which is a good sign for anyone if they are calling you back again. If not, don’t sit there, try somewhere else because they won’t tell you till end to leave. If you really want to get an interview, try to get in touch with organizer guy from company’s side.
3rd were these really cool dudes who started of with a matrix problem where each element was sum of row^2 and e^column. Now I had to prove something about equality of sum of elements that I don’t clearly remember but ultimately he would drive the result to generalization of function of row and column numbers.
The other one asked me a puzzle about two carts placed on a track with sensor below them. You can start cart with specific speed but afterwards you can control the motion only when cart passes sensor and when it leaves again you have no control over cart whatsoever. How to cause a collision between the two of them? They were quite helpful as I couldn’t quite articulate the problem properly. They helped me to get a hang of otherwise very simple problem (Talk to them all the time, they help you out a lot). They showed quite an interest in my low CPI (Prepare good and “genuine” reasons for it, they come really handy).
4th interviewers were really senior. They asked me about general stuff like my interns and interests. They were quite fascinated by my “Interest” in finance and “Global Markets” 😛 (Please don’t fake it, instead try to make a habit out of it, works like a charm in every interview if you get good at it). Then they asked me about Yelp dataset challenge and NLP featured in my AI project. They asked some basic questions about libraries featured in Scikit (Python Library on Machine Learning Techniques). Finally they asked me about flash crash in stock market related to another project in intern.
Then came a senior from my department who was in my hall too, working at GS. He asked me if I had any interest left in Electrical (Apparently that was my department for the whole time which had no role whatsoever this whole time, even in eligibility where it was open for all). He was really cool and talking to me in campus lingo. After that he asked me a question in data structures to which I knew best approach but I chose to start from higher complexity. Then “clicked” the lower complexity solution which was not right. Finally I came to an approach using tree which was not the solution he wanted but he couldn’t find flaw in my solution. Then the row sorted and column sorted matrix problem of finding an element came. Starting from highest complexity then lower and then lower. I told him all possible solution except the lowest complexity in which I demanded some hint because I “couldn’t think of any possible solution”. He helped me out and finally I told the solution starting from bottom left instead of top right (Acting recommended only for experts, don’t try if you are not 100% sure that you can fake it :P).
After this whole torture which went on till 11:30 am (they won’t let me go to any other company), they finally called me in and told me that they were “interested” and would like to “meet” me soon. Instructions were clear and then they asked me if I am going to appear for any other company to which I gave an “honest” answer.
If you like GeeksforGeeks and would like to contribute, you can also write an article and mail your article to firstname.lastname@example.org. See your article appearing on the GeeksforGeeks main page and help other Geeks.