Haowen is proud to be a Baynote scientist –  Principal Scientist that is. An employee for almost two years, Haowen really knows the ins and outs of our technology. We took a minute to learn about Haowen’s multiple areas of expertise at Baynote.


What’s your name and job title at Baynote?

Haowen Chan, Principal Scientist

How long have you been here?

1 year 9 months

What does a typical day at Baynote look like for you?

I wear a wide variety of hats at Baynote: R&D manager and facilitator, R&D contributor / implementer, software engineer and engineering coordinator for modernization. One example of what I do is that  I have recently been working with our Program Management Office to lay out how our different engineering projects flow together in order to meet  our modernization milestones for 2012. It’s an exciting time for the department because this is a big challenge and engineers love big challenges.  However,  we have to be very careful because we won’t be able to accomplish everything that we want in a single leap.  That said, we have to be sure that our individual technical priorities sync up coherently and make sense globally. Because the computation component of the platform is the essential core of the system which touches all the other parts, I help provide input to the technical design and prioritization for a large number of disparate pieces in our architecture.

Tell us how your past experience made you the right fit for Baynote:

My PhD degree specialization is in distributed systems security. Security in computer science is a cross-cutting discipline across many different fields including computer systems, networks, distributed algorithms, and number theory. You could say that I’m trained to be a generalist, and being a generalist allows me to comfortably speak the language of the diverse groups of experts I work with daily.

What attracted you to wanting to tackle the e-commerce personalization opportunity?

One of the most interesting applications of computing is how technology facilitates the generation and analysis of very large volumes of human-originated data. Dan Darnell, our VP of Product and Marketing, introduced me to the term “data exhaust” which I thought was very appropriate: data whose collection is almost incidental to the purposes of conducting the primary business. Now, the bigger the technology, the more data exhaust it generates, and the more powerful the technology needs to be to effectively make use of that larger exhaust. In this sense, the field has an interesting property that the technical problems basically continue to grow as quickly as, if not faster than, the technology. This ensures that we will always find new interesting and relevant problems in the space. Baynote is situated squarely at a nexus in this space with regard to personalization because we have an ongoing relationship with a large customer base with some of the most recognized names in e-commerce; this allows us unique opportunities to draw data, perform testing, and extend our applications suite in new and innovative ways.

In your opinion, what sets Baynote apart from competitors?

One of our many key differentiators is our ability to reach out and directly affect the behavior of end-users. For example, for product recommendations we can choose a product to show and see how we can adapt based on the response to that stimulus, rather than passively sit at the end of a data source and feed analytics to a third party indirectly. The ability to essentially engage the user in a dialog is invaluable.

Personalization is everywhere these days, from shopping to search. What’s a fun use of personalization in technology you’d like to see?

Maybe I’ve been watching too much Derren Brown, but I’d like to see how a system could be developed to essentially react to and affect a user’s state of mind. It’s an interesting problem because a large part of the conversation would be either unconscious or implicit. Current commercial products like Baynote can make the more immediate associations (people who buy this sad movie buy other sad movies) and many commercial artists use a lot of these techniques in advertising and in marketing, but I’m not aware of any automated way to deploy and leverage personalized mental effects at scale. Yes, a slightly less ambitious version of this problem is actually on the Baynote R&D roadmap, but  it might be a while before we get to it.

It’s Saturday. Where can we find you?

I like to go for walks in the Bay Area’s many beautiful nature parks. This is a truly beautiful part of the world with so many different micro-climates and ecosystems.