Tim Kraska

Within the Department of Computer Science at Brown University we are currently looking for strong students with system-building skills and out-of-the box thinking capabilities to develop the next generation of Big Data management systems. Brown's database group is one of the world-leading groups in systems-oriented database research, known for example through the H-Store, C-Store, SciDB, Aurora and Borealis systems. The ivy league university in general ranks amongst the top 20 places in the US for computer science and provides a nice work/life balance through its affordable living cost (important for PhD salaries), student culture, amazing restaurants, closeness to downtown Providence, as well as proximity to other bigger cities (Boston - 45min with the train, New York - 3h with bus or train).

If your are interested, please apply until December 15 at http://www.cs.brown.edu/grad/doctoral/.

• TupleWare: a new system for analytics, which combines ideas from databases, programming languages, and networking to revolutionize the way we do interactive data exploration. Initial results already demonstrated significant speedups compared to existing analytical frameworks (Hadoop, Spark, R, ...) and opened up a whole range of interesting research questions.
• New (secret) project, which will reduce cloud computing cost by an order of magnitude
• MLbase - A Data Management System for Distributed Machine-Learning
• MDCC - The Fastest Strong Consistent Multi-Data Center Replication Protocol
• CrowdDB - Integrating people as part of Big Data management
• PIQL - Performance Insightful Query Language
• and more top-secret ;-) ones

There are plenty of web pages out there on how to apply for grad school. They are still relevant and you should read them. In the following I will give some additional tips, in particular for students from abroad, on what you can do during your undergrad studies and for you application, if you want to work with us (the data management group) as a PhD student

When you apply:

• Timing: Start to prepare as early as possible. In particular getting letters will take a long time.
• Letters are extremely important. Try to get a recommendation from somebody we know. For example, if you (personally) know a professor who regularly publishes at VLDB, SIGMOD, ICDE, EDBT, etc., the chances are very high that we know him. Ask this person for a letter, as it will have a stronger weight for us than other letters.
• Name the person you want to work with in the letter
• Credibility: Put yourself in our shoes: we get hundreds of applications and it is very hard to evaluate them, in particular, if we do not know your letter writers or your school and grades compared to students from top-tier schools in the US. Thus, whatever you can do to make yourself comparable will help you in the end: list your grade ranking, MOOCs you did, or whatever else you can think of (see also next bullet point).
• Special project: In some cases, I provide a special project for candidates to solve. The project is usually a complex programming task, often with a research component but without requiring deep research knowledge. If you feel up for it, shoot me a mail and I will let you know. It will be time consuming and stressful. If you can solve it in a week, I will give a positive recommendation to the PhD committee. It is no guarantee that you get in, but it will increase your chances.
• Application fee waiver: If you have a degree from one of the top-places or and you come with the highest recommendations, but the application fee is a problem for you, please write me. Under certain circumstances we can waive the fee.

In case you read this year(s) before you apply, some tips on how you can work towards getting in our group:

• Do a research internship with us! Every year we offer 3-5 students the chance to intern with us. There are no time restrictions: you can do it in the spring, summer, fall, winter, spring and winter, and so on, and even apply short notice. If you intern with us and you do a good job, the chances are very high, that we accept you as a PhD student. However, getting one of these internships is almost as competitive as the PhD program.
• Massive Open Online Courses (MOOC): MOOCs as offered by Coursera, edX, Udacity, Canvas are a great way to make yourself "comparable" to other students from top-tier universities, in particular if you apply from abroad. If you rank high on one of these MOOCS with thousands of students, it can change the way how we evaluate your application.
• Try to do an internship at a large company, preferably in the US, before you apply. To apply for the PhD program, you need good recommendation letters and if somebody from IBM is saying you are an amazing student it is better than if a startup we never heard of is saying it.
• Try to work with somebody we know. For example, if a professor at your university regularly publishes at VLDB, SIGMOD, ICDE, etc., the chances are very high that we know him. Try to work with him as we will weight his recommendation letter higher than others when you apply.
• Try to be involved in a larger open-source software project. Essentially prove to us that you are an amazing coder. It is not a requirement (we also do a lot of theoretical work), but one of our key interests is building systems. Even better: you know how to really code in C++!

The American PhD system is very different from the one in Germany or Europe in general. In this document (in German) I try to explain some of the differences and how to best apply for the PhD program at Brown with a German Bachelor or Master degree.