Bitly has a lot of data, and our science team is on the front-lines working to make sense of this information and help answer questions about how we use the social web. This month’s employee profile highlights our Lead Scientist, Brian Eoff.
Brian hails from the south and never pictured himself living in NYC. He’s currently on hiatus from his Ph.D at Texas A&M and came to bitly for what was supposed to be a three month stint back in 2011, but has stayed on the past two years.
What are your main responsibilities as the Lead Scientist?
I help determine the research direction of the company, thinking about where our data can lead us, what new technology there is and approaches to presenting data in insightful ways. I give the science team space to explore ideas they find interesting.
What cool things are the Science Team working on?
We are currently working on using bitly for content discovery.
We are also taking some of the static analyses we have previously worked on and looking into making them real-time, which can be difficult. Visualizations work if you are sitting on their page and watching them change, but if you are coming back over the next few days, how can you see what has changed? How do you give historical context to it? If you come to the visualization once, how do you give the viewer the ability to explore more? We are working with the design and UX team to have a handle on that.
Can you talk a bit about the development of Realtime?
We were very interested in this idea of recency, and asked the question “what unique insight does bitly have?” We knew what people were clicking on at that time.
A lot of work went into this idea of “bursting” [content]. Certain things always pop up, like Obama and Justin Bieber. Everyday there are 30 articles about what Obama did that day. [We had to figure out] how to not always display that content, but [instead show] what is surprisingly getting a lot of attention today.
How did you get started at Bitly?
A little over two years ago I was doing my Ph.D with James Caverlee at Texas A&M. I didn’t want to spend summer in Texas, it’s quite hot, and I had also decided I didn’t necessarily want to go into academics. It was a moment to ask questions about where I wanted to go. I [made a list of] five places I wanted to work and wrote them each a letter, including [current scientist emeritus] Hilary Mason.
I was only supposed to be here 3-4 months, but my project was interesting and I wasn’t done with it. Hilary sat with my advisor and negotiated for my services. I still meet with my advisor every other week. I’ve been very blessed to know amazing people, him and Hilary are incredible and have really helped my career.
What have you learned working here?
A lot. I’ve gotten a lot faster at programming, taking ideas and turning them over. Hilary and [former bitly scientist] Mike Dewar made me better at the science approach. I’m surrounded by infrastructure and dev guys that can do amazing stuff very quickly and they just make you better at what you do. It’s a weird accomplishment when they tell me something i did was good, it’s like “yes!”
I’ve learned how to take something that is heavily mathematical and complicated and know it well enough to explain it to a diverse audience and relate what you do in a simplistic form. Some people worry that when you make something simpler you make it less precise, but I don’t think that. You can be precise but still be understandable.
You often speak at conferences, what have you talked about?
Normally it’s whatever I want. It’s sometimes on “What is data science?” or “What is big data?” Or it’s about what I’m interested in the month or two prior. I once talked about finding your worst enemy online. I never reuse material for talks, I like to make it new each time.
I previously gave talks in academia, but here you are talking to a broader audience, [and you have to think about] how [the talk] is perceived and how that affects the company.
Do you have any projects outside of bitly?
I advise Mahaya, a startup co-founded by Cornell professor Mor Naaman and Tarikh Korula, for a few hours each week. I met Mor when he was doing his Ph.D at Stanford.
I also do jujitsu. I always have to have something physical to do otherwise i get bored. When I came to New York I thought “what’s a sport I’ve always been interested in?” There are tons of amazing schools in Manhattan for jujitsu. Some of the best guys in the world have schools here. I’ve done a few tournaments while here. I try to train after work if I’m not too tired.3 months ago