neural networks

After my interviews for software engineer positions ended miserably1, I started a new journey to hunt for my first full-time job. This time, the goal was to establish stable income early as well as to learn more about machine learning. Therefore, I decided to interview for machine learning related positions.

WorldQuant – Quantitative Researcher

Since I had interviewed with Jane Street before1, I got interested in related industries. I’ve also heard that the salary at WorldQuant was quite high, and therefore I’ve decided to interview with them. Of course, I had heard about many different opinions about this company, but I thought it would still be an interesting journey.

The first stage of the interviews was a mathematical exam. I came across my classmates as I attended the exam and we encouraged each other before leaving. Luckily, I was able to pass the exam.

The on-site interviews involved problems that had a mathematical flavour. It took some time for me to come up with a solution. Rather than encouraging us to speak out our thoughts like in interviews elsewhere, the interviewer stopped me when I tried to explain my thoughts early. He told me to think alone first, and started using his mobile phone. Although I was a little bit uncertain about the situation, I was able to come up with a solution at the end.

When I asked the reason why he chose this company, the interviewer was quite honest. “It was all about money. Some people might want to publish papers about investment. I think this is quite useless. It only matters when you are able to earn money.”

“Perhaps your could still publish some of your methods after it’s no longer effective in the market?” I asked.

“But publishing is not important for me.”

“Just like I said, I only care about earning money.” The interviewer responded.

After the interviews ended, I went back to the school, and started to thought about it. Indeed money was really important, but I still felt that I was unable to like such ideas. Therefore, I decided to quit the interviewing process.

Zillians – AI Developer

I came across the recruitment ad for a AI Developer position at Zillians on the Internet, and I felt very interested. Therefore, I applied for the job.

Because at that time, I felt my background on artificial intelligence was not very strong, I decided to write a detailed letter explaining my background and motivations. I really don’t know why I would think that would be useful, but luckily I did get invited for the interviews.

Before the interviews, I was given a homework to analyse source code of Caffe framework, and to make some modifications. I spent quite some time on it, and attended the interviews. The interviews involved the usual programming exercises as well as questions about the Caffe code.

I asked the interviewer why he chose Zillians.

“Why did you choose Zillians among other choices?” I asked.

“Because Zillians was such a rare company. If I did not go to Zillians, I would be afraid that it might go bankrupt very soon then I would not be able to work at such a company anymore.”

“I could always to go work for other companies later because most of them would still exist by then.”

The company featured a open space office just like other startups, but it was also possible to work alone quietly if you like. As we talked, the interviewer brought me to a room where only one employee was working inside. I said hello to the lonely employee before I left the company.

Afterwards, I had second rounds of interviews via Skype. The offer I got was very competitive, but I decided to turn down the offer.

Appier – Machine Learning Scientist

Appier was a very well-respected company in machine learning and I also had a friend who worked there. Therefore, I asked my friend for a referral.

I was asked some machine learning problems as well as programming problems during the first round of interviews. The second round of interviews were similar but also involved questions about my masters thesis.

I still remember asking interviewers questions like “Why did you decide to start up a company?” It was indeed an enjoyable chat.

I successfully got an offer.

Facebook - Machine Learning Engineer

The interviews with Facebook happened in a different time from other interviews described in this article. However, it’s convenient to put it here.

The story started when a recruiter from Facebook sent a letter to me via LinkedIn, offering me about opportunities in machine learning. However, because I was too tired from day-to-day work, I hadn’t replied it for at least a month. I then got a second letter, saying that it was the last chance to interview.

Some of my friends also talked about the email about the last chance. “Looks like I am not the only one who is hesitant to reply. " I thought.

I decided to try anyway, and surprisingly I passed the phone interview without any preparation, unlike my last attempt1. Perhaps the bar for machine learning engineers was lower? Or perhaps it was helpful that I still practice problem solving sometimes during these years? Or perhaps I just got lucky?

But the on-site interviews did not turn out positive. I performed poorly on distributed machine learning system design and basic machine learning problems. I felt sad, because it appeared that my machine learning skills actually did not improve a lot during these years.

On the other hand, I also felt relieved. Because if I did get the offer, my career path would be largely decided because I might be too afraid to try different paths due to financial considerations. Perhaps this was an opportunity for me to take a detour, to see what could be possible ahead.