Internships at Microsoft Research

What are internships at Microsoft Research like?

Are you a PhD student interested in my research (or any of the other researchers in the BioML team at Microsoft Research?) Consider applying for an paid 12-week internship with us. While industry internships are more common in computer science, I’ve noticed that a lot of biologists and computational biology students aren’t aware of these kinds of opportunities. Thus, I wanted to write a blog post explaining how internships with the BioML team at Microsoft Research work, because I genuinely believe they can be life-changing in broadening students’ research and career trajectories, and would love to see these opportunities extended to a more diverse range of students across the life sciences.

MSR is likely more similar to an academic lab than you expect. Researchers largely have free reign over their projects and research agenda and often there are not immediate ties to existing business or products at Microsoft. Researchers are primarily evaluated on academic impact and scholarly outputs (e.g. publications), and many choose to work on basic science. For this reason, the internships at MSR sometimes look different from what many students imagine when they think of an “industry internship”. For example, instead of being handed a pre-packaged project that can be seamlessly integrated into business operations, internships at MSR tend to be more flexible and “academic” in scope. Interns often work together with their mentors to define projects based upon their shared research interests, and many end up publishing the work they do here, or incorporating it into their thesis.

With this context, I often describe internships at MSR as an opportunity to work for 12 weeks under a different academic advisor (i.e. one of the researchers at MSR). This can be a highly impactful opportunity (and different students often take different impacts, depending on where they are in their research and career), but some possible take-aways from an internship include:

  • Broadening your research interests and skills: Researchers at MSR work on varied research areas, and often with cutting-edge techniques: working with them is a valuable way to learn new skills, topics, and methods. Some of my interns come from biology backgrounds and are interested in learning more about machine learning; other interns come from computer science PhDs, and are interested in learning more about biology.
  • Exposure to interdisciplinary work and collaborations: MSR is incredibly interdisciplinary. For example, MSR New England has researchers in computational biology, social science and human computer interaction, machine learning, economics, healthcare, mathematics, all working on the same floor. It’s impossible not to meet people outside of your research field, and you’re encouraged to sit down with researchers and attend seminars from all over the academic map. These kinds of interactions can really broaden your awareness of research beyond your immediate bubble, and spur new collaborations.
  • Seeing how research works in a new (industry) environment: Research jobs are incredibly diverse. It can be hard to see this if you’ve only worked at one lab your entire PhD: seeing how research works at MSR can help broaden your perspective on what you might want to do after your PhD (and many of our interns tell us it has!)
  • Access to compute resources: Since there’s the potential for more compute resources at MSR than what most PhD students would have access to in an academic lab, many interns do take advantage of this to launch larger-scale projects.

If any of this interests you, keep reading for information and tips on how to apply.

(In addition to our PhD internships, we also offer undergraduate internships, which are also research focused. I won’t go into them in detail, but a lot of the information I share here is still relevant - although of course, we don’t expect undergraduates to have as much prior research experience and tend to scaffold projects more.)

How do I get an internship with the BioML team at MSR?

Generally, we put out internship applications in the late fall. You can monitor our job posting website, but many of the researchers will also put out calls for applications on social media (e.g. Twitter). In general, an application package will include an academic CV and letters of reference: we use these materials to understand your research profile, your interests, and what projects you’ve been working on up to date. We’re also likely to ask for a research statement in this year’s cycle.

Some tips and advice about the process:

In general, the researchers lead every step of the internship selection process, from reading your CV, to selecting and interviewing candidates, to finalizing offers. It can be worth it to pay attention to which lab and topic area the internship is hosted in (e.g. you’d be working with different researchers if you applied for a computational biology internship at Microsoft Research New England versus Redmond), and browsing through the profiles and websites of the researchers at the lab to understand the research they do. This helps you better envision how your research fits in with ours.

The process is driven by individual researchers, and we all have different criterion in what it is we’re looking for in interns. The most common aspect is topic area - we usually want to find interns that have at least some overlap with our own research interests. But in addition to topic, there are many competing and overlapping considerations that will shift from year to year: driving certain research agendas forward over others, using the internship for mentorship and impact on students, exploring or expanding new interests, etc. There are two main takeaways from this. First, there’s no single “ideal” candidate we’re looking for. For example, our intern cohort this year was incredibly diverse: some were early PhD students, and others were late in their PhD. Some came from biology programs, others came from computer science programs. Some had a more established research portfolio and lots of completed projects, while others were in more of an exploratory phase of their career. Second, if you apply and don’t get an internship with us, it’s much more likely us than you. The majority of CVs that we read are “above the bar” and are from candidates who would benefit from an internship with us. Often, these candidates are filtered out for reasons outside of their control (e.g. we might be interested in pursuing a different research direction that year than the candidate’s expressed interests.) I mention this because I’ve heard of applicants being hesistant to apply to MSR because they’re early in their career or don’t feel accomplished enough (e.g. they don’t have enough papers) - don’t take yourself out of the running early, because you simply don’t know what the researchers might be looking for that particular year.

We usually receive a lot of applications, and have a limited amount of time to review each package. While we do our best to synthesize your research profile and interests from your CV, a website with a summary of your research interests (or even a paragraph about this at the top of your CV) can help. This is especially helpful if you have a more complex background (e.g. research papers in many different topics or your current research interests differ from your previous published papers or if you changed fields midway through your career) - this lets you frame your research, instead of us having to (potentially incorrectly) infer it. Second, if your research synergy or aims are less obvious, consider introducing yourself to us over email and/or asking for a video call to chat. Often times, our research agendas aren’t fixed in stone: if a candidate comes from a different research area, but has some solid ideas of how they might bridge their research area with one of ours, this is one example of an internship candidate that could be very fruitful but might be overlooked from the CV alone. A change this year compared to previous years is that we’ll ask for a summary of your research and aims for the internship, to try to make sure everyone has a chance to communicate this information: I would suggest you try to think about the above points in drafting your statement.

Finally, if you get invited for an interview, expect us to focus on your research. In the BioML team, the interviews often look like unstructured “research chats”, and many interns are surprised that we don’t require a coding exam (or usually drill deep into technical aspects.) Instead, think of it this way: we chose your application because we were interested in the research you’ve listed on your CV. The interview is now an opportunity for us to see if you can explain the research. Good questions to think about include: what are the big picture problems and motivations underlying the research? How did that inform the questions and approaches? What was your role in the research project? What do you want to do next, and what kind of role do you see us playing in mentoring that?