A Summer at Microsoft Research

It's autumn now, a time of harvest and reflection, and the beginning of the academic year. The blog has been dormant for about a month because I've been working very hard in Astro-land.

I spent the last few months only 10 miles away from UW, just across the lake in Redmond. Since lots of people have been interested in how the experience was, and since corporate internships seem to be fairly uncommon in Astronomy, I thought it would be worthwhile recapping my summer at Microsoft Research. (Apologies for a super lengthy post. tl;dr MSR was fun, challenging, would recommend)

How it Started

The whole summer adventure began with a tweet. Seriously, I landed a summer internship because of Twitter. I was scheduled to give an Ignite-style talk at the Hard Rock Cafe in downtown Seattle for a Vis conference and saw this tweet via someone I "follow":

Hey, PhD students into #bigdata and #datavis: Msft Rsrch accepting intern app for Summer '13. Contact me!
— Danyel Fisher (@FisherDanyel) November 29, 2012

I ping'd him, since I didn't have any summer funding and my Starbucks article had just "blown up" a bit, making the rounds on some legit news and TV. Danyel, a super cool guy, was going to be at the Hard Rock. I didn't horribly embarrass myself, and after my talk Danyel agreed to meet me for coffee (so Seattle) to casually chat about data, academics, and internships. Some emails and a Skype interview with some of the team followed, and in late February I got the offer from Microsoft Research (MSR).

Stay in academia? Leave for the corporate world after graduation? This is a fundamental decision grad students are wrestling with in my field, where funding is shaky and career prospects are uncertain. I wanted more data on how the "other half" lives, and happily accepted.

During the cafe chat, Skype interview, and throughout the summer it was very clear that this blog was a big reason they offered me the internship. My advice to other grad students who might want to explore the corporate world is not to write a blog (unless that's something you enjoy doing). Rather, I suggest finding your way to build an exterior network. Do something you enjoy that engages that fancy brain you've spent so long investing in, and that has some broad appeal. I use the blog to ask silly questions that interest me, share mostly via Twitter, and have attracted some press attention using social/news aggregation sites (reddit, digg, slashdot, etc).

Arriving at Microsoft

Even though I grew up in WA, and spent most of the last 10 years in Seattle, I had never been to Microsoft. First impressions: the Microsoft campus is beautiful. It feels sort of like a university, lots of native plants, big buildings. There are lots of people my age (late 20's - early 30's).

The orientation seemed pretty standard. At Microsoft when you start a job or internship they say you are "on-boarding". If NASA is king of jargon and acronyms, MS must be in a close second. Lots of standard first day stuff, photo for my name-badge, coffee and snacks, a few hours of boilerplate instructions/introductions.

"Any questions?"

One kid in the back sheepishly asks "What about using... technology... from other companies?"

The MSR researcher fielding the Q/A replies "Security, take this man away...." and we laugh. He produces an iPhone 5, smiles, and then says "I wouldn't worry about it". It's a silly question, but gets at the heart of the culture. None of us knew what the answer would be in advance.

By lunch I've met up with my mentor (boss), Rob, our first time meeting in person. I'm relieved that he's easy going, and we chat about coffee, cooking, and eventually the summer plan. Walking around MS/MSR I quickly find that I'm a stranger in a strange land, as the saying goes. The locals are friendly, and there's loads of fun things going on for interns, but (as you might expect) most people have a CS background.

Life as an Intern

I worked for/with some very neat people. As far as I could tell everyone at MSR is brilliant. Industry in no way appears boring or intellectually vacuous, as the rumors/implications in astro-land would have you believe. Here's some of the team I worked for.

Here is the "official line" on what the life of an intern is like in a blog post and a cute video. I think these are pretty accurate.

All hail he free coffee machine.

It's immersive, hands on, fast paced, and generally exciting to be there. Academics rejoice! There is frequent free food; I had lunch provided at least once a week, either in lunch-talks or team meetings! Every couple weeks there are drinks/snacks after work. Ample parking, day or night. A city bus pass is provided, as well as access to Microsoft's extensive shuttle-bus system. There is free soda, juice, and coffee. No joke, you can have as much as you want... and I did!

The pay is good, competitive for CS internships. By Astro-Land standards you are apparently richer than King Solomon... for 3 months.

MSR is a special place. It's basically an enormous CS department. Nearly everyone has a PhD, all the interns are PhD students (unlike normal interns, who are mostly college students). Every single day there are as many as a half-dozen visiting speakers, often faculty at premier universities, giving talks on huge range of CS/tech topics. A few notable ones I attended: an author of a D&D book, the CTO of Weta Digital, a data scientist at a leading social network company.

Boys Named Sue: A visualization I made one day while at MSR

Curiosity and fun is encouraged, at least from interns. Some people played board games at lunch. About a half dozen of the interns (incl. myself) made faux research posters for an art exhibit that hung in the MSR lobby for a month. Mine was titled "An Extensible Pataphysics Engine", and drew heavily from Dante. Within MSR I found you're largely shielded from any of the "evil MS" culture you may have heard about. These things may exist in, e.g., product teams, but I don't know. I never felt like I was being picked apart by crows.

Building 99 - Home of MSR in Redmond, WA.

The research team and my mentor were all very open and available. In a given week I would go to ~3 group meetings. I worked normal hours, coming in around 830, leaving around 6 most days. Lots of the interns had similar schedules. The team I worked for studies user interaction with big data (here's a publicly available publication on their research). There were 3 interns on our team, each of us working on our own projects. I met with Rob and/or Danyel every day, and we chatted about data, visualization, design... and sometimes just about life. My intern project was largely testing the interaction with data within the software they were building (to wit, I was a "captive data scientist", as Rob put it).

This was a fairly unstructured charge, and while I'm a pretty self-sufficient worker in my academic life, I stumbled a bit at first to find what my "job" was at MSR on a daily basis. "What would I produce at the end of the summer?" I wondered. This uncertainty, combined with the task of learning new lingo and new programming skills, caused some serious anxiety. In the first month I came home several nights and asked my (amazingly supportive) wife "what the hell am I doing with my summer?!"

After about 4 or 5 weeks I got over the angst/growing pains, and focused in on a single data exploration. This more directed project was my idea entirely, and using data I would only have access to while at MSR. Rob/Danyel were immediately supportive and encouraging, which was critical since I was reaching as an astronomer trying to do research that CS/HCI people would care about. This culminated in a research paper, a "note" that I submitted to a top level CS conference on my very last day of work! Hopefully it will get accepted, and then I can publish a blog post/link to it! This paper will definitely be going on my CV, and I consider it the product of a successful summer of research/learning. I think my mentors would agree

Closing Thoughts

Seen on my last day at MSR.

If I were to do it again I would push harder initially for finite/actionable ideas on projects. Having rock solid focus on what you want to produce is critical for a successful internship. Its clear that the gang at MSR would have supported my wanting to do almost any creative research project in our niche. I would absolutely do another internship, at MS/MSR or another tech company.

I would also recommend tech/industry internships for other astro grad students. You could work on data, design, analysis, machine learning, visualization, or HCI. I learned a ton, and even though it was 0% related to my thesis research, it was a healthy and productive experience. I (try to) seek out broad research experiences in astronomy, working/learning on a wide range of topics, because I love the synthesis that results from combining them. My summer at MSR fits nicely in that model.

MSR and MS are not the faceless mega-corps of your nightmares, but are run (from what I could see) by very nice, charitable, honest, hardworking people. They are trying to make good products, research new technologies, that solve a vast range of problems. However, MS also has many people who believe that unlimited supplies of Mt Dew is a valuable perquisite. Take from that what you will.

Industry now seems more approachable, but simultaneously more daunting. You don't simply "leave academia" and stroll in to the land of milk and honey, welcomed by trumpets and showers of money. These people are smart, creative, and hard working. TANSTAAFL

For now I'm excited to get back to at UW, with under 2 years left in my PhD in Astro-Land. So much research to do! I don't know where I'm going next (really, who does?) but this summer I've grown in some interesting new ways.