HOW TO BE SUCCESSFUL INDUSTRY RESEARCHER – A journey from NUS-HCI Lab to Adobe – Part I

“AVEID” Paper Accepted to ACM IUI 2018
December 12, 2017
HOW TO BE SUCCESSFUL INDUSTRY RESEARCHER – A journey from NUS-HCI Lab to Adobe – Part II
January 4, 2018

HOW TO BE SUCCESSFUL INDUSTRY RESEARCHER – A journey from NUS-HCI Lab to Adobe – Part I

Rubaiat Habib is currently working at Adobe Research. We got the opportunity to share his success story starting at a research lab and living his dream of working at Adobe.

So we have Rubaiat with us. Thank you for sparing your time. Why don’t you go ahead and introduce yourself?

I joined NUS-HCI Lab around 2009, and started working with Prof. Zhao. Time flies! I graduated in 2014. During my Ph.D. time, I was fortunate to do internships in Microsoft Research, Japan Science & Technology Agency, and Autodesk Research. Later I joined Autodesk as a researcher. In addition to Prof. Zhao’s supervision, these internship experiences were invaluable for my research career.

Keeping it general, what do you think are the two most important things all students in academic research should know?

I’ll just talk about three high-level things I think which are very important as a researcher. First of all,

“What do you really want to do?”

What makes you happy? What are you passionate about? What motivates you? This is a hard question. Even I don’t have all the answers for myself. But it’s really important to ask this question to yourself. Once you find out what excites you, the next question is

“How can you be the best at that work?”

There are many ways of preparing to be best in a field of work. When I joined HCI Lab, I remember that we used to have this weekly CHI reading group. We used to read all the papers in the CHI collectively as a group. This gave us an overview of entire HCI field. Everyone got to know exactly what’s happening in HCI. It is extremely important to do a lot of readings. Not just papers but tech magazines, books, social media and so on. The second really important exercise we had at that time in HCI Lab (I don’t know whether you do it now), is writing. Prof. Zhao made us write one paragraph every day. He (Prof. Zhao) did this exercise to improve our English skills. But in retrospect, I think that it not only improved our writing skills, but also improved our thinking skills. I think, in order to be strong researcher, you have to write a lot because writing is a thinking tool. Most of my research ideas came from reading books.

I also found that a lot of students lack communication skills. You should talk to people when you go to the conference, talk to people as much as you can. As an Asian student, I was always shy when I went to a conference and it’s always difficult to talk with random people when they are talking to someone else. You don’t know how to start a conversation. But there are other ways of doing it. You can find the researchers, students and professors working in the same area and send them e-mails before the conference. Just schedule a meeting during the coffee breaks and so on. All people in the conference want to talk and interact.

Let’s go to individual questions.

After working in the industry for several years, whether your view about research changed?

Yeah. It changed. The first thing I learned is that publication is an important aspect of your contribution, but not everything. I mean for your PhD, publication is really important.

When I look at my colleagues, the most successful groups or individuals are those who have a very strong, consistent story/vision towards their research, along with tangible outcomes.

How is publishing in academics different from industry?

In industry, usually, your impact to the organization is a combination of things. Publications is just one of them. You know, everyone is trying to innovate. But, judging innovation is not easy. Publication in top-tier conferences give credential to your innovation. But, at the end of the day, in industry, the practical impact of your work is what matters most. So, publications are important, but, probably not as much as in the universities.

How to be successful in the industry?

Think about bigger problems, learn and adapt continuously, and collaborate with great minds!
Think of what your career should look like 10 years later, and what do you need to do towards achieving that goal!

Having big impact is a combination of things – great papers and patents, product outcomes, influencing product roadmaps, sometimes you want to make a new field and invite more people to contribute to that field.

Focus on the quality of your work. When I got my job, I only had three papers. One or two really good papers can make a big difference in your career. I saw people getting great faculty jobs but have only two or three papers. The other takeaway is:

To do great projects, you have to collaborate with the great people. It is challenging to solve everything or have big impact all by yourself. You will have to work with great people, have great team members, communicate with the students & colleagues so that more people are passionate to do what you are doing and help you out towards your mission.

Our field (IT) is changing so rapidly. It’s so diverse and ubiquitous. To do great works, you have to bring many disciplines all together.

So I think it’s really important to keep that in mind. How to be a great collaborator is a very important skill. I know many great researchers; they are outstanding, fantastic individual researchers. They are very hard working, very talented; they have all of technical knowledge. But they didn’t really reach their full potential because they are so focused on their own work, not really open to collaboration where they could have invited more people to help them out for a bigger outcome. I think these are really important skills: being a good team member and inspire other people to work with you.

Great paper/project, great story (motivation), great people.

People, who do extremely well in the field in their early careers, are all great collaborators. They have a lot of conversation; they are working with a lot of people. So that’s something to keep in mind.

How is research different from product development?

First of all, in product developments, you are focusing on a particular product. Your domain and scope of work is often narrower. Whereas, in research you have a much more holistic approach. You are not working on a particular product; you are working on long term strategies and vision.

How to be a successful industry researcher?

The people who can lead long-term vision and strategies for the company are the most successful researchers/leaders.

In research, there are a few things that are used to evaluate your success— papers, patent, product (technology transfer), press, people (influence on people in the field/company).

Other than conducting research, researchers can have a number of roles in an industrial lab – mentoring students/postdocs, thought leadership, technology transfer (productization of an idea), laying out the company strategy on a particular field, influencing product teams, and creating certain cultures in the organization (e.g., Bill Buxton).

One particular example is the Draco, published in 2014. But in 2015, 70 percent of my time was dedicated to the tech transfer to design, deploying it to the beta, inviting people to the beta, managing the beta community, running contest for feeding artworks. That was a significant amount of time and effort. But it was totally worth it!

Success Story of Draco at AutoDesk

When I look back, Draco was my most favorite project. I have other favorite papers, but it is certainly my favorite project due to its final impact and outcome. When the research transfer team took the concept and developed the prototype, the product teams were not ready at that time to take on this project. Their business goal was not directly compatible with Draco. The Autodesk consumer group was closing down; they were just killing the existing app, so they were really not interested in taking the new app. But what really helped was the fact that we told a great story. We managed to convince them that Draco was something beyond animated drawings. It’s a powerful tool for visual communication, which could affect a wide range of people. We invited a lot of people into the beta forum, and they were very excited about the product. Draco was just the beginning of a very exciting future; there is a big need for this kind of media managing. It’s not for artists; it’s for educators, for teachers; it has a broader range of potential people and it will give unprecedented edge. So when they saw all the people creating great concepts on a very private community, beta community, and when they saw architects, designers, teachers using it. They were sold. After that they took the prototype and released the product. When I look back, the original research paper is only a small fraction of it. We had to do one year of development work and then we had to invite this beta community, we had to do a lot of marketing internally and externally to the different product teams to handle. Why is it an important direction? It is because a company doesn’t want to increase their product portfolio. They have to maintain and market it. It’s a lot of commitment, so they will only do a new product when they see the significant value it brings. All these things made huge difference. Without this beta form, without this active involvement and without the fact that I was involved so much with the beta people, ran those contests, it would not have been a product. So those things made a huge impact. For one year I was just doing all these things. And later once it was released, I focused more on research.

How much academic freedom does one have in Industry Research?

It depends on the team and the company. Some labs and groups enjoy a lot of academic freedom (i.e., publishing), whereas, some other groups are more secretive. Either way, if you’re working for a company, usually, you’re expected to do something that serves the company’s business and purpose.

How does Publishable/Non-Publishable Research Impact Your Career?

There are many ways of doing research. Even in Autodesk, some groups are focused on academic publishing, while other are focused on more advanced prototype development. There are many other exciting research groups that we are not even aware of. I know a friend of mine who is working on the Google Autonomous project and I know he’s super excited about it. They don’t publish but they are also working on really cutting-edge and hard problems which are definitely worth publishing. But Google and Facebook and Uber they’re not publishing.

Publishing is not their main focus. It’s just potential career path.

Apple, Amazon, Facebook, Uber, Google; they’re working on really big ambitious problems that will have huge global impact in all of our life. There are a lot of great people working in those areas. These days, a lot of tenured professors are joining these companies because of the problems they’re working. They have the data, they have the engineering resource, and they have the expertise. Even if you work all the same you might not get as much academic exposure but you will have a great career, because

The type of work you do make you unique.

 

To BE CONTINUED: you will find out Rubaiat’s view on “What kind of skillset is required to be successful in the industry?”, “how to make most out of your internship opportunities?”, “The main take-away points” and much more in the second part of the interview. Stay tuned!

Leave a Reply

Your email address will not be published. Required fields are marked *

%d bloggers like this: