Each summer, you can hear those at Microsoft Research expressing enthusiastic gratitude for the annual influx of interns who swell the ranks of labs worldwide.

It’s easy to understand why. These young researchers from around the world bring a certain mixture of enthusiasm, expertise, energy, and recent academic achievement.

This year, those sighs of gratitude must number at least 895. That’s the number of interns Microsoft Research is hosting at its six labs. The researchers who mentor the incoming interns assign challenging assignments tailored to the skills and interests of their new charges.

Rest assured that most of those mentors also get a genuine thrill at the prospect of collaborating with new, up-and-coming colleagues. And in many cases, these are professional relationships and friendships that will endure for years.

A few of this summer’s Microsoft Research interns discuss the opportunity—as do the researchers who get the pleasure of working alongside them:

ARASH ASADPOUR

Arash Asadpour

• Hometown: Qazvin, Iran.
• University: Stanford University.
• Internship:iMicrosoft Research New England, July 1-Sept. 30.
• Why I chose Microsoft Research: Why not? Great researchers, ambitious post-docs, very motivated interns, a vast number of visitors in various relevant and/or interesting areas, everybody interested in collaboration, a very friendly environment, a beautiful location, and so on.

Every single one of the above might be tempting enough to convince me to spend a summer at a research center, although I have to admit that before I started my internship last summer in Redmond, I was only aware of the first two or three items.

• What I’ve been doing: Last summer, I worked on algorithmic aspects of resource-allocation problems. This time, I am mostly involved in the game-theoretic issues. In particular, I am working on two projects. One is to study the dynamic of congestion games and its effect on the efficiency of the game. The other is to define and analyze some relevant—and, hopefully, useful—variations of the stable-marriage problem, which has been used to match such things as medical residents with hospitals and kidney donors with recipients.
• My experience: This year, when I found out about the newly opened New England lab, I was really excited. In addition to all my other reasons for an internship at Microsoft Research, the lab is more focused on my areas of research interest. It is just next door to MIT and, almost, Harvard. And Boston is fantastic!
• Mentor says: “Arash is a fantastic intern working at the boundary of algorithms and economics,” says Jennifer Chayes, managing director of Microsoft Research New England. “In order to come up with these algorithms, he draws on his strong mathematical background, as well as his impressive intuition in economics. He’s already done some significant work on fair allocation of indivisible goods; he obtained some of these results while he was an intern in Redmond last summer. This summer, he’s engaging with our economics visitors and interns, as well as algorithms people in the lab, working on a variety of problems in algorithmic game theory. His new projects include some very relevant modifications of what is known as the stable-marriage problem.”

GABRIELA CRETU

Gabriela Cretu

• Hometown: Sinaia, Romania.
• University: Columbia University.
• Internship:iMicrosoft Research Silicon Valley, June 2-Aug. 22.
• Why I chose Microsoft Research: As a Ph.D. student, it is very important to get the feeling of the research industry so you know the options at the end of the program. In the Ph.D. student community, a Microsoft internship is considered one of the best experiences. I was excited to have the opportunity to experience it myself.
• What I’ve been doing: I have been working on analyzing the logs generated by the DryadLINQ platform and devising techniques that can lead to the detection of anomalies in the system, based on the logging information.
• My experience: The internship started very seriously, and it came with a nice learning process. The DryadLINQ platform has amazing properties, and learning about it and then analyzing it is very interesting. The enthusiasm of my two mentors contributes a lot to the great experience, and to have everything, meeting everybody in the lab, and participating in social events with the other interns make for an excellent experience.
• Mentor says: “Gabriela’s project involves building a tool for analyzing the behavior of large computer clusters, such as the ones used to process data for search,” says Mihai Budiu, a researcher at the Silicon Valley lab. “The analysis itself is performed using a parallel programming language developed here, which is executed on the cluster itself. This enables the tool to analyze large quantities of data efficiently.

“Gabriela’s work has two parts: build the basic data-gathering tool, and design and implement statistical analyses to mine the data. These analyses have already enabled us to diagnose some performance anomalies occurring intermittently in a 250-machine cluster.”

SANJAM GARG

Sanjam Garg

• Hometown: Patiala, India.
• University: Indian Institute of Technology Delhi (will be attending the University of Texas at Austin to pursue a Ph.D.
• Internship:iMicrosoft Research India, June 16-July 16.
• Why I chose Microsoft Research: I wanted to work on the theoretical side of security, so working with the Cryptography, Security, and Applied Mathematics group at Microsoft Research India looked like an ideal option. Working on problems in provable security sounded very challenging and interesting. I also wanted to see research being performed in a corporate research-lab environment.
• What I’ve been doing: Researchers try to argue the security of cryptographic schemes they construct based on the computational hardness of well-known problems. They base their belief of unbreakability of a scheme by constructing an efficient way to solve a hard problem assuming existence of a party that can break the given scheme, technically called a reduction from the hard problem to the given scheme. The existence of such a reduction rules out the existence of a party breaking the scheme and, hence, is a proof of security of the given scheme. The efficiency of the reduction in these cases is very important.

In our work, we analyzed the Schnorr-signature scheme, a discrete log based signature scheme that is one of the most efficient signature schemes known. Unfortunately, the known reductions from the discrete log problem to the Schnorr-signature scheme are not very efficient. Our contribution has been in establishing that a very efficient reduction can’t exist, based on the fact that existence of such a reduction violates other computational assumptions.

Such impossibility results reveal a lot of interesting details about the structure of the schemes being studied and are very interesting from a mathematical viewpoint. From the viewpoint of application, such negative results are very important, as they help quantify the exact level of security that can be reached used a particular scheme.

• My experience: Lots of fun. The facilities and the working environment are great. I got an opportunity to interact with a number of senior researchers and was exposed to the state of the art in lots of diverse areas of computer science and allied areas. I also made a lot of friends. I really value the professional exposure I got while I was here.

[Mentors] Raghav Bhaskar and Satya Lokam are both amazing researchers, and it was great to work with them. I learned a lot more than just about signatures.

• Mentor says: “The focus of Sanjam’s work was understanding how tightly the security of certain digital-signature schemes can be related to the hardness of the discrete log problem,” says Lokam, a researcher at Microsoft Research India. “Sanjam continued to work on this as his B.Tech project after getting back to IIT Delhi [after his 2007 internship at Microsoft Research India]. His tenacity paid off, and we were able to make significant new contributions to research this topic. In particular, we improved the lower bounds on the inevitable gap between the computational resources needed to forge some digital signature schemes and to solve the discrete log problem. We believe Sanjam will continue to be a valuable collaborator with our group as he takes on more and more challenging problems. Our experience working with him has been a pleasure.”

DONGGUN KIM

Donggun Kim

• Hometown: Seoul, South Korea.
• University: Yonsei University (South Korea).
• Internship: Six months at Microsoft Research Asia and July 8-Sept. 26 at Microsoft Research Redmond.
• Why I chose Microsoft Research: I think it is very good experience to have an internship in a global company like Microsoft. I noticed that there was a special internship program supported by the Korea Research Foundation, consisting of six month at Microsoft Research Asia and, optionally, three months at Microsoft Research Redmond.
• What I’ve been doing: I’m working on a project to make JavaScript (including Ajax) applications faster using “speculation.”
• My experience: At Microsoft Research Asia, I worked on the Point2Share project, to make device pairing and file sharing between mobile phones easy, intuitive, and secure. That work won the Best Demo Award at MobiSys 2008.
• Mentor says: “This summer, Donggun will be working on a new idea that could make Web-based applications much faster,” says Jeremy Elson, a Microsoft Research Redmond researcher. “Imagine having your e-mail appear the moment you click on it, map views update the moment you tell them to, and online calculators give you answers almost before you ask. Using a technique known as “speculation,” we hope to make this a reality. Like an attentive butler, future Web sites may someday start working on your answer even before you’ve finished asking the question.”

OLGA KULYK

Olga Kulyk

• Hometown: Kiev, Ukraine.
• University: University of Twente (the Netherlands).
• Internship: Microsoft Research Redmond, May 13-Aug. 1.
• Why I chose Microsoft Research: I got inspired by a great talk by Mary Czerwinski [research-area manager at Microsoft Research Redmond] at the Technical University of Eindhoven, the Netherlands. I was at the beginning of my Ph.D. at that time, working on evaluation of collaboration support for large displays. I knew right away that I wanted to do an internship with the VIBE [Visualization and Interaction for Business and Entertainment] group at Microsoft Research. I just knew I would love Microsoft Research’s dynamic, inspiring research environment.
• What I’ve been doing: I have been working on the design and evaluation of large display visualizations to support shared awareness of development teams’ activities and project health in collocated team spaces. I worked on this project with Mikkel Jakobsen, Mary Czerwinski, Kori Quinn, Roland Fernandez, George Robertson, and Gina Venolia.
• My experience: These past months at Microsoft Research have been just terrific. I got to work with a great team and got to meet a lot of amazingly talented and very fun people! I have learned a lot from my colleagues, other human-computer-interaction interns, and the development teams we’ve been designing the visualizations for. Everyone at Microsoft has been incredibly helpful assisting me in my project. I enjoyed living in the Seattle area, hanging out with cool researchers and interns whom I will miss a lot after leaving. I am very grateful to my colleagues at Microsoft Research, my recruitment team, and all the great people at Microsoft for this awesome experience!
• Mentor says: “Olga came to intern with the VIBE team because of her background in group scientific work utilizing multiple display surfaces, including large, touch-screen displays,” Czerwinski says. “She was keen to help us evaluate our second version of FASTDash, a large display visualization of a dev team’s activity in a code base. Together with many others on our team, including Mikkel Jakobsen, she is deploying the technology in two Microsoft dev teams’ environments and will spend several days logging, coding, and observing the usefulness and usability of the tool.”

MICHAEL PEDERSEN

Michael Pedersen

• Hometown: Ringkøbing, Denmark.
• University: University of Edinburgh.
• Internship:iMicrosoft Research Cambridge, June 16-Sept. 5.
• Why I chose Microsoft Research: An internship is a great opportunity to get a fresh and alternative perspective on my Ph.D. work and meet potential future collaborators, and there are some great people at Microsoft Research. The choice for my internship was easy: To my knowledge, it is the only commercial research lab that does basic research so closely related to my own research area, in computational biology. Microsoft Research is also a good candidate for future job opportunities, and an internship experience will help me decide if I fit in here. Oh, and then, there is the money …
• What I’ve been doing: I am working on a programming language for synthetic biology, together with Andrew Phillips, my Microsoft Research supervisor [and a scientist at Microsoft Research Cambridge]. Synthetic biology aims at designing and constructing new biological systems. Some popular examples include cells that produce energy or detect and report environmental pollutants. There are many languages around for describing existing biological systems with the aim of simulating or analyzing them, which is great for understanding how biology works.

We are designing a language with the aim of programming new biological systems and translating programs to actual DNA sequences. The ultimate dream is to program one or more cells in some future version of Visual Studio, press a button, watch a computer translate the program to physical DNA, and putting this to work in lots of living cells.

• My experience: I spent the first few weeks of my internship exploring different options for a research project and had a high degree of freedom to learn, to explore various ideas, and to pursue the ones that interested me the most. I also had my first-ever biology wet-lab experience putting together real DNA and playing with real cells, which was enormously exciting for an otherwise theoretically minded guy like me. This was organized as part of the University of Cambridge iGEM [international Genetically Engineered Machines] competition—a worldwide annual competition where teams of undergraduate students compete to design cool synthetic biological systems—in which Microsoft Research is involved.

On the social side, a number of cool events have been organized for the interns here, including gaming and movie nights, a boat trip in the rain, and a day trip to London, although I haven’t been able to take part in all of them. I am particularly excited about an upcoming go-carting event for the interns and supervisors!

• Mentor says: “Michael started his internship with a clear idea in mind: He wanted to build a programming language and compiler for generating a genetically modified organism from program code,” Phillips says. “After attending some courses on synthetic biology, organized by the University of Cambridge iGEM program, Michael set about scouring the literature for interesting case studies of genetically modified bacteria. He collected a range of systems, from pollutant detectors to predator-prey bacteria engaged in deadly warfare. Now, he is busy designing a language for programming the behavior of these systems and figuring out how to compile the programs to a collection of DNA sequences from a registry of standard biological parts, or BioBricks. The challenge is twofold: On one hand, we need to come up with the right language, and on the other, we need to extend the BioBricks registry itself in order to better model the properties of its components so the right parts can be automatically selected during compilation. If Michael has his way, genetic engineering of bacteria is about to become a whole lot easier.”