Sumedh Joshi bio photo

Sumedh Joshi

Applied Mathematician

Email Twitter Facebook LinkedIn Instagram Github

Just FYI, these tips are adapted from a talk I gave to the acoustics graduate program at The University of Texas.



In 2012 I won the National Defense Science and Engineering Graduate (NDSEG) Fellowship to study the interaction between shoaling nonlinear internal waves and ocean acoustics. To be perfectly honest, I had never considered myself a student that was worthy of such awards, and didn’t expect to win for several reasons.


First, my grades were okay, but not great (something like a 3.5 out of 4.0). Second, the NDSEG requires that an applicant be in the first or second year of graduate study. I was in my third year, but had some extenuating circumstances. Namely, I had just changed institutions and majors. Finally, and perhaps most critically, I was going to apply for the fellowship without a letter of recommendation from my research advisor, who (at the time) I had just started working with and didn’t know very well. Basically, I applied for the fellowship because I had an idea of what I wanted to work on, and it was not that much work. I didn’t apply because I thought I had a good shot of winning. However, having successfully won the fellowship under these circumstances, I think I have some useful advice to offer.

NDSEG information

  • The NDSEG has about a 5-10% acceptance rate, which is strongly dependent on the area you apply for. They used to release statistics on the acceptance rate by application area, but I don’t think they do anymore.

  • It pays $30k, $30.5k, and $31k in three consecutive years. It, unlike the NSF fellowship, cannot be deferred. Also, depending on where you go to graduate school, it may pay less than a graduate research assistantship. At Cornell, I took a paycut to take the NDSEG.

  • It is sponsored by the Army Research Office, Office of Naval Research, Air Force Office of Scientific Research, and the High-Performance Computing Modernization Office (my sponsor). As such, you should be thinking of one of these institutions when you write you application letter.

  • It has a ridiculously easy application process when compared to the Hertz or the NSF-GRFP. The NSF requires several pages of essays, and the Hertz requires an in-person interview if you make it that far. The NDSEG only requires a 3000-word essay.

Application tips

  • Everyone should apply. Technically it’s supposed to be in your first year of graduate study barring extenuating circumstances. I applied in my third year of study but had enough extenuating circumstances (new school, field, advisor, and project) that I was awarded the fellowship.

  • Brainstorm a project; this is on you to come up with something interesting and fun. Don’t worry too much about having to adhere strictly to it, but make sure it is something you are actually interested in doing so that that sincerity comes across in the application. If you do not have such a project in mind, seek out mentors in your research group to help you brainstorm (which is good advice for an academic in general!).

  • Write your essay early, have someone who has written grants to the funding agency you are applying to read it. Have other NDSEG winners read it. Have lots of people read it and give feedback.

  • Reference the literature in your essay, at least by names of people who might be good collaborators. They don’t have to be at Texas. These are also great people to have read your essay to judge whether your research plan is feasible and a good idea.

  • Use the jargon of your field to make sure your essay gets read by people who know your project. If there are recognizable institutions you are affiliated with (ARL) reference those to make sure that its clear that you’re a contributing member of the research community.

  • Write your essay like a mini-grant proposal: that is what it is.

  • Get letters of rec from people who are either well-known in the field, know you really well, or are likely to write really well. Ideally they’re academics or research scientists, and ideally they satisfy more than one of those tenants.

  • Read Philip Guo’s page for general advice on fellowship applications.

  • Grades and scores are not critically important, my grades and scores were both below average for graduate students. However, make sure you demonstrate that you have a good idea that you are sincerely interested in.

Departing thoughts

In general, you should apply for a fellowship when you feel ready to formulate a problem that you are interested in working on. In my opinion, it is unfortunate that most of the big fellowships are only awarded in years one and two of a doctoral degree; it was a little bit later that I was more certain of what I wanted to do. But, you should use these applications as good opportunities to solidify for yourself what it is you want to do. If you can describe cogently and clearly what your plan for graduate school is, you’re probably a good candidate for the NDSEG, and, other fellowships.