My (Evolving) Intern Selection Strategy

This is part of a series of articles on internships. Previous posts talked about the motivations of doing and offering internships (Why Intern?) and my motivations for taking on interns (Why I Work With Interns). This one is about how I go about deciding whether you are the right candidate.

First a bit of background — I’ve been fortunate to go through the Shell assessment training program, where we went through the philosophy of a good assessment and how to go about making the right decisions in a structured, objective, non-judgemental, unbiased manner. This helped me to formulate necessary and sufficient characteristics of good interns and how to evaluate these qualities in candidates.

Unfortunately, trying to do this in an interview is too time-consuming, hazardous, and ends up being imprecise. Basically, a 1-hour conversation is insufficient, and multiplying this by the number of candidates is inefficient. 

In order to rectify this, I tried doing a ‘pre-internship’, where I invited students to work with me and then we would decide if we wanted to continue. This didn’t work out too well for a number of reasons that I shall not go into here.

My current approach is to create a bunch of filters to make things easier for me and students interested in an internship.

Filter 1: Technical capability

I ask students to work on a small problem based on their area of interest. This is not a pass/fail test! I am more interested in understanding how students went about researching a new topic, developed an understanding of the problem, and developed a solution and how they solved or tried to implement the solution. Thus far, I’ve asked students to work on:

  • digit recognition using vector distances
  • AGM visualization of across a range of (pairs of) numbers
  • simulate and visualize the behavior of an RC circuit to a step input

They get a week to work on a problem, and I encourage them to consult google, Wikipedia, etc. but to implement their solution themselves. I don’t think these are challenging, and an internship would need (at a minimum) the ability to solve these problems. 

There has been a spectrum of responses:

  • Some students don’t respond — I guess they find this too difficult, or it is out of their interest
  • Some do work on it, but struggle, and realize that maybe this isn’t for them
  • Some of the outcomes are not up to snuff, so I gently suggest that they work on their basic skills and re-apply
  • Some don’t solve these but demonstrate the ability to independently pursue new areas and come up with innovative approaches
  • Some do a fantastic job, and we have great conversations around their ideas and implementations

Note that, once again, solving a problem or not does not guarantee or disqualify you from an internship. 

The only thing I don’t like about this is that it is a ‘negative’ filter that removes candidates, but so be it! This has saved me untold hours, and has resulted in a few interns that I’m excited to work with!

Filter 2: GitHub

This has not been too useful. My goal is to understand if you have been working consistently on any area and to take a peek at your code. Unfortunately, students don’t have the discipline of using a repository; many don’t even have a github account! 

In addition to technical capability, this also helps surface other qualities that I am looking for. More when I write about this in a future post.

Filter 3: Your CV

Your resume says a lot about you:

  • Are there any simple grammatical/spelling errors? It’s perfectly fine to not be proficient in English, but there are so many tools that highlight these. If you cannot take care of the document that represents you, I don’t have much confidence that you will take care of a document that represents your project!
  • Is it formatted well? I’m not looking for artistic designs, but I do expect a minimal level of consistency
  • What are the things that you are most proud of? And do these matter to a successful internship?

Filter 4: Your email/message

I do not use this as a filter, but including it here as food for thought.

Unsolicited emails convey a lot in terms of how you introduce yourself, and how you express your goals, interests, and motivations. 

Responses to presentations/LinkedIn posts/Articles are fascinating. I explicitly ask for certain pieces of information, and most respond appropriately. Others just say ‘Interested.’ This signals that you are unable to follow simple requests, and can be a pretty powerful filter as well. 

Share your thoughts on this approach! If you are a student who has gone through this process, I’m very keen on understanding your perspective.


Leave a Reply

Your email address will not be published.