Round 1 Interview with directors
The interview was really short, only 30 mins. Beforehand, I was told that the interview is going to be a mix of behavioural and technical, but from rumours, I was told that if it was with directors, its likely that it will be behavioural. However, that was not the case
When I entered the interview room, I was told to introduced myself. Before I could complete my self-introduction, I was rudely interrupted by one of the older directors. He asked me if I know what DSTA is doing. I mentioned that I was previously in ISG, and the job description seems very similar to what I was doing last time in ISG. He asked me to then explain how I can implement AI in what I was doing during my NS. As I was told that it was most likely behavioural, I explained in a very high level as to what the AI can do. The director mentioned that he expected way more technical depth and reminded me of the data structures and CS fundamentals that I learnt. I tried to explain in more technical depth, but he did not seem happy. One of the younger directors then cut in and tried to help me asking me to explain in 3 different steps (which I forgot what the 3 steps was oops). I then explained the whole data pipeline. How its currently being done, so we make use of excel sheets, and we need some process to extract out those data and perform some form of data cleaning then represent those data in a database schema, and I explained an example of a schema. Then I mentioned that we can then feed these data into our AI model for training purposes. This AI model can then also be used to detect and generate image identifications, which then can go through a validation process (by human) and feed right back in to train the model. The younger director that mentioned that I got 2 out of the 3 points that he asked me to explain (oops…).
The younger director that asked me some resume deep dive questions. I don’t remember all the questions he asked me, but here are some memorable ones. He saw my CS2103T project, then asked me what is the differences between a desktop app and a web app. Interesting question, I mentioned that in a desktop app, we have more isolation and control over the libraries we use as all the files that is needed to run the program will be in that executable. Not sure what he was looking out for by asking this question though… He also saw my ChargeEV project and asked me how I would scale it. Think I answered this well, though maybe I could have done a little better. I mentioned that I could separate the backend to different micro-services and deploy them in kubernetes, so when one service has more traffic or requires more power, we can scale horizontally by spawning more pods. We can then slap a kafka between the microservices and the database, so we can scale horizontally, and for the backend, we can continue using noSQL to have fast reads and writes. (I feel that my answer was probably enough for them, but there are still some issues, as I failed to mention how kafka integrates with the database)
After the interview, I asked the interviewers what is the tech stack they use. The older director was enraged by my question, and mentioned “Why do all of y’all keep asking what tech stack we use? So if we use a certain tech stack, then y’all don’t want to work anymore? It just shows that y’all are not flexible enough!” I tried to defend myself to say that the tech stack of a company says a lot about a company, such as Ascenda using Ruby means they prefer simplicity and elegance. (very interesting director lol…)
Round 2 OA
omg hella hard, thank god for chatgpt… Interestingly, I did not pass one of the correctness test case for the first question, and had a bunch of TLEs. Second question had one TLE.
Somehow got offered tho, lol