Graduate School Admission Tips from the Inside.

This is that time of the year where graduate school admissions are upon us with me being far from ready to take that leap; although I have been nursing the dream of one day getting admitted for a Masters in Computer Science and then a Ph.D, I still think I need to immerse myself more in the industry and of course earn additional money along the way.

I came across this very insightful and candid post by Prof. Amin Vahdat, a CS professor at UC San Diego after stumbling across his blog from a Slashdot link that talks about Facebook's really impressive architecture to handle 300+ million users and counting.

This lead me to check the UCSD CNS web site where I noticed that Matt Welsh, now a Harvard faculty, (a name I vaguely remember to be related to NBIO and SEDA) would be giving a talk in the CNS Lecture Series come January 2010. He also maintains a blog: http://matt-welsh.blogspot.com/ where he also has a post on How to get into grad school for intending graduate students.

I particularly enjoyed the witty and simply hilarious post he linked to by Prof. Luis von Ahn, Computer Science professor at Carnegie Mellon University. My best part:

7. DO mention the name of a professor that you want to work with, but make sure the professor is still alive.

Now that definitely is the type of advisor I'd like to work with, one that wields a seemly boundless sense of humor.

At least both Profs. Amin Vahdat and Matt Welsh are active in the research areas I am interested in: operating systems, distributed computing, concurrency and computer networks but the real challenge is if I can blast through the GRE and bring my research skills up to snuff to earn glowing recommendation letters that are important in getting me into their schools. I've never been a stellar student (although I once came top of my class in nursery school :D) so I'll have to rely on doing insane stuff when I go back into academia for a MSc. I'm thinking of throwing an MSc. in Economics into the mix since I also have a real strong knack for business but haven't quite nailed how I can apply this. I'm vaguely thinking applying distributed computing on some business problem that currently takes requires a lot of resources to complete similar to this but for a truly novel business problem.