On My Experience Applying to International PhD Programs (as a Student Who Studied in Morocco)

The mind is that which is roused and directed by itself. It makes of itself what it chooses. It makes what it chooses of its own experience. — Marcus Aurelius, Meditations

Making the decision to do a PhD can be daunting, and knowing what it takes to build a strong application can be even more opaque. Throughout the process, I had to spend hours looking for the right information, especially since finding someone with a similar background was almost impossible. Later on, I was also contacted by many students struggling with the same questions, a sign that maybe my questions were not so out of place after all, and that clear answers to them were still hard to find.

So I decided that writing such an article might not be out of place (or useless) after all. So here it is. Bear with me as I try to cover some of the things I wish I had known three years ago.

In retrospect, I think the decision to pursue a PhD (and hence make my CV and work experience research-oriented) was one of the best decisions I have made. But again, you can only know what something is by knowing what it is not, and you can only know whether you and research get along well if you know from experience that you don’t picture yourself working in a non-research-oriented industry.

Disclaimer: The following is based on my experience as an Élève Ingénieur in the Moroccan (French-inspired) engineering system. While some details may be specific to this system, the general advice should apply regardless of your academic background.

Now, why do I think that a PhD is worthwhile? Simply because I’m someone who looks for a greater meaning in what I do on a daily basis, and I couldn’t picture myself spending years doing something that didn’t fit this description. A PhD was one possible — and, in my case, the only — path towards such a career.

You should ask yourself the same question: Why do you want to do a PhD? To help you answer it, I’ll try to convince you in the following section not to do one :)

Why you might not want to do a PhD

• Doing research doesn’t necessarily entail having or doing a PhD: in fact, some of the most influential researchers in the ML field don’t have a PhD (e.g., Alec Radford, one of the main figures behind the GPT architecture, and Chris Olah, one of the pioneers of mechanistic interpretability. Olah also wrote this piece that you might find useful, since we’re talking about whether university degrees are that important).
• A lot of the breakthroughs in current ML come from researchers working in industry: this is partly because they are more comfortable budget and compute-wise. They naturally end up being able to take on some bigger questions and experiment at a scale that can be difficult to achieve elsewhere.
• There are fewer PhD spots, and not necessarily good offers for you: especially if you’re coming out of an engineering school, given the salary you’d start with and the raises you might be subject to within less than four years.
• A PhD is mentally hard: you need to figure out your research direction, fail multiple times, question yourself, and sometimes question whether what you have been working on for the past six months is worth it.
• You might think about quitting… And you might think about it more than once.
• Life instability: you might think about the stability you could have had if it weren’t for the PhD you chose. You might be far away from your family, living in another country, adapting to a completely different environment, and dealing with all sorts of uncertainties that you probably never had to think about before.

But if you survived this far, and if the idea of doing a PhD still excites you, congratulations! A PhD might indeed be a good path forward for you.

Now, given the tiny chance of landing your dream offer, let’s try to maximize the competitiveness of your application and minimize the risk of you ending up with something that falls short of your expectations.

Building your profile

I would say this phase should start in the 3rd year of your engineering school (or the 1st year if you’re coming from Les Classes Prépas).

My first piece of advice is: PLEASE TAKE CARE OF YOUR GRADES! You can’t imagine the number of people who regret not having pushed themselves a bit more back in school. The reason is that all of the PhD programs I applied to paid close attention to my grades (and some awarded me a scholarship based mainly on my GPA). The heuristic here is to aim for As everywhere, so that you have some margin for error and minimize the risk of your GPA falling below the “safe” range. This is no joke, since PhD committees often use grades as an easy criterion to screen out candidates.

Moreover, please maintain a good relationship with your professors. Ideally, ask if you can join their lab and help with an ongoing project. That way, you can secure a strong recommendation letter. But if that’s not possible—and it isn’t in a lot of cases at engineering schools in Morocco—then at least you have tried :)

And this brings me to my 3rd point. If your professors don’t offer Research Assistant (RA) positions, don’t give up (never surrender, actually, but we’ll get to that in a bit). Try reaching out to other professors instead. At my previous university (UM6P), there were a few professors on the research-focused side of the university who kindly accepted me as an RA. This can strengthen your CV enormously.

But if RA positions are simply not possible for you, hackathons can be a good backup plan. Join hackathons, build projects, and develop your portfolio. You never know which project might eventually help you land your first RA position. For instance, one of my hackathon projects helped me tremendously during the interview that led to my first international RA position at NUS.

RA positions play a huge role in the pre-PhD phase. First of all, they introduce you to research and help you figure out whether you really like it. They enable you to develop expertise in a given topic, which might eventually lead to a publication. And finally, you’re paving the way towards securing a strong, research-oriented Letter of Recommendation (LoR). Bingo!

By the way, your professor here can later become your PhD supervisor (or at least one of the professors who might offer you a PhD position).

Finally, I would also like to mention this: it’s rare for a major AI conference to be held in Morocco (Interestingly, this year conferences like EACL and AISTATS have been held in Morocco). But when it happens, be sure to seize the opportunity. Conferences are a great way to discover the latest trends directly from the research community, meet and exchange ideas with researchers, and, importantly, meet professors who are recruiting students. You won’t have to pay for anything if you apply for financial aid or student volunteering.

Preparing the application

As mentioned before, the general rule of thumb is that you need to showcase that you’re a good and reliable student. This often shows up through your transcripts… so take care of them.

At this point, I can also assume that you have some prior research experience and, ideally, one or two conference papers. I know it seems like a lot…and it is. But, as with everything else, the PhD market has become increasingly competitive. If you want to secure a well-funded PhD in a strong research environment, you’ll find yourself competing with people who graduated from well-known universities (Stanford, Harvard, Tsinghua… etc.), have multiple papers accepted at top venues, and have near-perfect GPAs.

Given this, professors will often want to minimize their risk. Taking on someone who has already done research at a high level (and therefore knows what research actually entails) and still wants to pursue a PhD is simply less risky for them.

So perhaps the safest thing you can do is give professors the green flag they’re looking for. Again, we’re not discussing whether this is right or wrong. We’re trying to get you into the PhD position of your dreams.

Your LoRs: if you have done two or three RA experiences, then you’re covered. The remaining LoR can come from one of the professors who taught you a course relevant to the program you’re applying to.

Finally, your Statement of Purpose (SoP). This one is considered the masterpiece of your application, and I’d say spend a lot of time thinking it through, refining it, and getting it read and double-checked by other people.

I’ll share some examples of SoPs here, including my ETH one.

Basically, it’s in your SoP that you tell them what kind of researcher you are. It’s where you explain the thread that links all the work you have done and the work you intend to do during your PhD. And it’s where you explain why that particular PhD position and institution are a good fit for you.

It’s one of the most important pieces of your application. Treat it as such.

Emails, interviews, and waiting

Prior to officially applying through university portals, it’s always a good idea to check the websites of the professors you’d like to work with and see whether you find a lab or research group that would be a good fit. You can then contact the professors by email. Not all of them will answer (which is not necessarily a bad sign, by the way), but some of them will.

If they do, it’s a great opportunity to ask for a Zoom call, introduce yourself, ask your questions, and see whether there’s a good fit. Some professors already have a good idea of who they’re going to recruit well before applications officially close, and they often recruit people they have already worked with as RAs or people they have interviewed (you, hopefully).

Prepare for interviews: They are stressful. You can be asked a math question, a ML question, or even to code (rarely, and this seems to happen more often for positions in Europe). You will also be asked about your previous research, so be ready to explain it clearly and motivate the choices you made. Finally, you will likely be asked about your future research interests. Have a good idea of what these are, even if you know they may evolve later on.

Rest and wait for the answers: Most programs will get back to you well before April 15, although the exact timeline varies by country and university. And while checking Reddit for updates is often tempting (and fun, to be honest), if it’s adding more stress than it’s worth, simply try to forget about it and get yourself busy with something else.

A personal note

Believe it or not, I decided to stay in Morocco precisely to pursue this unconventional path: to see whether I could build a strong research career without following the more conventional route of studying abroad. Thankfully, it didn’t turn out to be a complete failure! And now, I’m able to write this kind of article in the hope that it might motivate you to consider the same path.

I have often been told that studying abroad is the way to a good research career, and while there is some truth to it, it’s not the whole truth. You can study in Morocco and still achieve great things.

Of course, I won’t advise against studying abroad. If I were to make the choice again, I would probably still go for it (with some changes to the plan and destinations)... But just know that, whatever you choose to do, you are the one who decides where the path takes you and whether you make the most of it.