How to ask a potential supervisor for a (challenge) project?

Question

I intend to apply for master's positions in Computer Science and I have done some projects, especially in ML/AI/RL. I would like to ask for some guidance and even a challenge project, from a professor (or professors) who might be potential supervisors. I would like this challenge to showcase my abilities to them. I have done some projects but I think I can do better, if I have some guidance. the challenging aspect means that the professor could be more certain of my abilities in this area of expertise, and that it would be highly relevant to them, since there are many projects that are challenging, but the supervisor would not find them relevant enough.

How should I phrase such an Email? Should I include my CV and credentials? I am at a loss to find the right words to write, since this has never been done before. I would appreciate any suggestions or guidance. Thanks in advance.

Note: I am mostly targeting the US and Canada.

Answer 1

I wouldn't recommend making such a request at all, for the following reasons.

1. PhD admissions in the US (and I believe also Canada) is a departmental matter. Attracting the interest of a particular potential supervisor, while it does still carry some weight, is not decisive in the same way that it is in the European system.

2. Related to the previous point, departments need to make sure that the admissions process is fair to all applicants. If the decision to admit you were to be based on your performance on a project that was offered to you, while other applicants were not given such an opportunity, that would be a problem.

3. Coming up with projects for students that are challenging yet feasible is hard work. Speaking as a professor, I might be willing to put in the effort to create such a project for a current student, but not for someone who is merely a prospective student.

4. While there may be some difference of opinion on this, a student's ability to succeed on short-term, artificially-posed "challenge" projects may not be a good indication of their potential for success at research. Real research problems are much slower, messier, and less well-defined. They often don't have such clear goals or endpoints, and those endpoints can shift as the project goes on. It can happen that students who are great at "challenge"-type problems find themselves more frustrated by problems that are less neat and tidy.

5. Also, real research will involve interaction between the student, the advisor, and possibly other collaborators or members of the lab. A purely individual project like you propose will not provide any measure of how you work in more collaborative settings.

I think that if you write such an email, no matter how carefully you phrase it, there is at least a 99% chance that the recipient will decline, and will just ask you to apply to the program through the normal process.

If you really want the ability to demonstrate your abilities in project form, I would instead suggest finding a professor at your current institution to work with. That person can then write a letter of recommendation that will tell the committee about the quality of your work and skills. I think that is a better way to go, even if it the topic is not 100% "relevant" to what another potential supervisor works on.