Dr. Ying Li, Chief Scientist and Co-Founder of Concurix Corporation
“The Bellevue-Seattle region has a lot of experience in data mining… the Chinese market has not yet matured to the level we have here. I want to help with this.”
~ Dr. Ying Li
Dr. Ying Li is the Chief Scientist and co-founder of Concurix Corporation. Prior to joining Concurix, Dr. Li worked at Microsoft as Partner Engineering Manager and worked closely with adCenter and Microsoft Research to help develop company products and businesses. She helped establish data mining services and has over 14 years of experience in the field of data mining.
Dr. Li recently received the 2012 ACM SIGKKD (Special Interest Group on Knowledge Discovery and Data Mining) Service Award, for individuals who have made significant contributions to the research community, mentored teams, and helped build and advance technologies. Each year it grants two awards in innovation and service. These are considered the top international awards in the field.
Q: Why did you choose to move to Bellevue in 1998?
A: I had job offers in both the Los Angeles and Bellevue areas and chose Bellevue for its environment and sense of cultural friendliness. I’ve always liked cities and Bellevue has a metropolitan feel without being too hectic.
Q: You are the Chief Scientist and co-founder of Concurix Corporation. Can you please tell us about your career path to your current position and some of the work you have done?
A: I have always characterized myself as an applied scientist. After I earned my Ph.D., I went to work as an applied researcher at a government industry research center in Canada.
After that I joined Microsoft as a program manager, in close collaboration with Microsoft Research. The first idea I had was to establish data mining services into products, rather than the usual approach of program managers, in which development and software engineers go straight to a feature.
During my time working with Microsoft Research, we focused on applied research. It wasn’t curiosity-driven science. We were always working on the product side with a clear goal. With the product in mind, you contribute by solving problems with advanced algorithms. So a developer would solve a problem through coding, and my contributions would be through algorithms, experimentation and data manipulation. In combination with the developers, we would solve the problem together.
Q: You worked at Microsoft for many years and had a very successful career. Why were you interested in pursuing Concurix?
A: There is a commonly accepted understanding that it takes about 3-5 years to build a body of knowledge. During my 14 years at Microsoft, I had mastered three big topics to introduce new things for the company.
But after three cycles, I began to think of innovation in new fields and new methods of execution. Therefore, I had the desire to do a start-up.
The start-up environment and the Microsoft environment are quite different. There is some of the same spirit, in that you think outside of the box. But with a start-up, you have to execute outside the box and be prepared in everything you do. I think that was an important reason why I left. I wanted to break my comfort zone, and challenge myself.
The main reason I was drawn to this start-up was because of the subject matter. I wanted to see how we could improve operating systems through the use of data mining.
Q: As a start-up company, what does Concurix hope to offer?
A: If you look back 30 years from today, you can see that operating systems have been making a slow evolution. Hardware has advanced hugely; more and more computers are built with multi-core CPUs which provide powerful computing. In fact, on my second day at Concurix, I assembled a computer with a 64-core processor.
On the hardware side, computing capability has grown rapidly, but current operating systems are unable to fully utilize that power. We see an opportunity to address the gap. There is existing research and experimentation that demonstrates that traditional operating system capabilities will flatline or decrease at a large number of cores. When a computer has more than 16 core processors, performance decreases because the operating system spends more time managing the multiple cores.
We want to reexamine all the operating system designs and determine which one can maximize this potential power. We hope to give the industry a 10 times price performance gain. So for the same cost, you should be able to gain 10 times better computing performance than traditional operating systems provide.
Q: How do you like working at Concurix? Have you noticed any differences working for a big company as opposed to a start-up?
A: We just started in May 2012 so we are still in the development process. I do all types of work there. I self-write code, analyze statistics, and make calculations, but I also make coffee and answer phones.
Of course there are pros and cons to working in a smaller company. Bigger companies are organized and accommodating and have many resources.
As a start-up, we must think things through carefully and always consider our budget. I really enjoy the freedom we have, though. The decision-making cycle is much shorter so it is easier to eliminate the constraints that can delay development.
Q: Do you see any opportunities for yourself and Concurix in China or another Asian market? Are there any projects that you hope to start that will help further these opportunities?
A: We want our work to be global. China has had such huge growth of their computing centers and there is definitely potential for growth there. We have been talking to Chinese companies to establish partnerships and find potential clients.
The Bellevue-Seattle region has a lot of experience in data mining; we’ve been working with big data for the past two decades. We’ve got interested people from many companies and a wealth of resources.
Data-driven decision making started and is growing much quicker in our region than in China. In a sense, we can say that the Chinese market has not yet matured to the level we have here. Given my personal experience over the last 14 years, I feel as if I am among the earlier group of individuals who have been working on this topic. I want to help China with this.
When I was working for Microsoft adCenter, Microsoft Research established a team in Beijing. In the adCenter Labs I created with my boss and Dr. Harry Shum, we worked to solve adCenter algorithmic questions. We used Microsoft Research expertise and leveraged their development power to solve problems.
It was a very real experience. I led this team and helped to train these employees in China. We worked together for almost eight years. It was only a small-scale project at that time but I think I want to create this type of dedicated training effort. Between our wealth of resources and some of my own experience, we can find some way to help foster data mining in China.
Q: You mentioned that there is a good environment here in the Bellevue and Greater Seattle area. What do you think about the start-up atmosphere here?
A: Bellevue has established a good culture and environment for start-up opportunities. It’s an area with a lot of great potential. During the past 30 years, it has really benefited from the presence of companies like Boeing and Microsoft and institutions like the University of Washington.
We have a community where employees are challenged and encouraged to solve tough industry problems. Start-ups here are built with a genuine interest and desire to make lasting impacts on the region.
Q. How can we reach you?
A. You can contact us at firstname.lastname@example.org, 425.576.0300, or visit our headquarters at 244 Market St, Kirkland WA 98033.