Interview Experiences

The last few months haven't exactly been exciting, I've failed to get through numerous interviews. It kind of reminds me of the campus placements in December 2007, when I was shortlisted almost everywhere, and the conversion came very late, after almost 10 companies. Yeah, it was more baffling then, as I continued to belie expectations. I was perceived as the unfortunate hero of a tragedy movie back then, with my credentials I was supposed to get a good job easily. Easier said than done though, an interview requires extensive preparation, something I am not very fond of.

What exaggerates the problem in my case is perhaps the unwillingness to learn from mistakes, the interviewer's excitement at seeing my CSE background and my complete lack of skills and knowledge in coding and computer science in general. So it does seem ironic to me when I am approached for fundaes, but the satisfaction of preaching always supersedes any thoughts of guilt, hypocrisy or shame. This time around, I go one better, I post them online and publicly, and I definitely can't deny an anticipation of a dramatic increase in the number of unfortunate souls who come to me for doubts about how to go about interviews. In conformity with my repeated failures at wooing perspective employers, this is more of a 'my mistakes' list.

- When you are applying in an Investment Bank, you'd rather know finance basics.
I had cleared the first round convincingly, playing the AIR-15 card the best I could. In the second roud, I was asked about my knowledge in finance. I had obviously planned an answer, which focussed on my interest in the field (supported by nothing on my resume, not one course or certification or project) and my ability to learn things fast as and when the need arises. As it is, I sounded ridiculously stupid; I needed to understand that my competitors had taken harmless NCFM exams, many had done internships, et al. While my claim of quick understanding was not without proof, my competitors were no idiots themselves.

- While you should definitely know whatever you did in a project, it is also important to know about the larger picture.
Almost invariably, the interview focusses on the projects that one has done. In a recent interview, I was asked about my internship at INRIA Futurs, France. Now, as I started explaining my algorithm for propagating updates in XAMs, I was interrupted, asked about what exactly these XAMs were. I explained and the natural question followed, where will they find their use. I tried to explain the best I could, to the maximum possible extent of my fantasies where they could be employed, but my oblivion to the larger picture was easily evident, and my efforts in developing a new algorithm to reduce the time taken considerably were easily ignored.
In the same interview however, what I think was revolting and utterly idiotic on the part of the interviewer was his suggestion to backtest the data provided by the IT team in my company, who have been there for many more years than I was, and questioning me about the algorithms being used by them. Perhaps even more irritating was his inability to digest the fact that most data mining applications require manual exploration as a basis.

- Among two topics A and B, if you have read more about B, you should stress on the same.
Having devoted a good chunk of my preparation time to C++, it was an obvious folly on my part to suggest that my programming language preference was Java. I used to be very good at Java, but lets face it, I had not written a Java code for an year.
But then, there are things you just can't help, like when people get to asking syntaxes which they might have remembered for some random reason, but 99 out of 100 people copy paste the code using the technological revolution called 'Google'. Now, I am not sure if I should generalize this, but the tendency of Indian interviewers to do so is far greater than that of foreigners; I guess the Indian interviewer is more likely to act like that elusive 100th person. More generally, I believe that it is a discussion with interviewers from on-shore desks, and a syntax-based "Kaun Banega Crorepati" with Indians. It is quite possible that the latter appears so pathetic to me only because of my inability to remember syntaxes.

- When an interviewer wants you to make a guess, he usually wants an educated guess, and not some random value.
I learnt this the hard way during my campus placements. The overall guesstimate was a classic - to estimate the cost of a software that returns abstracts of all Web documents, assuming it is the only software that does so, and it is absolutely essential to get the abstracts of all documents that exist. To calculate the man-hours saved, you need to know the number of words that an average human mind can process per unit of time. On being asked to make a guess, I said 100. The interviewer shot back that it was a gross underestimation, and he could easily read and understand a page in around a minute. I think that it was true even for me, and it was silly of me to actually get to that value. At that time, however, I was too rude to accept my mistake and we had a good argument, not the best thing to do in an interview. He met me later that night, after they had taken the people they wanted, and rightly told me that I needed to improve on my attitude, a gesture that I now find extremely gratifying considering it was me who started the heated exchange of words.

- Prepare a good justification for the weaknesses in your profile
Invariably, I am grilled about my low CPI, regretfully in my case it is coupled with the thought of what I could have done, the only place where the AIR-15 tag makes life difficult for me. Some interviewers are satisfied with my explanations, many infer that I did not study at all.

I guess, this should be enough for now, I shall furthur enlist my experiences if the readers express such a request. Adios.