Employee Spotlight –
Lauren Burcaw, Lead Data Scientist

September 8, 2019


Lauren Burcaw is an award-winning physics PhD with nearly ten years experience in data analysis and modeling. A graduate student of one of the top nuclear magnetic resonance research labs in the world, she has also held research positions at NYU School of Medicine and Schlumberger Oilfield Services and is an alumna of the Insight Data Science program. She has numerous first author publications in prestigious academic journals and conference publications. At OpenSlate, Lauren has spent the last four years using her data science expertise to advance our capabilities in areas including brand safety, quality metrics, and contextual categorization.

OpenSlate spoke to Lauren Burcaw about her current position and role in the company, her team, future projects, and gave us insight into the company’s culture through the eyes of a Senior Data Scientist. 

 

What kind of personality traits thrive in your department as opposed to others? 

To be successful on the data science team at OpenSlate, you need to be curious, analytical, and collaborative.  We’re a small company so it’s imperative you do not work in a silo and since we’re working on cutting edge problems that may not have been solved before, you need to be unafraid to forge a new data path.  Additionally, you must be willing to get your hands dirty and familiarize yourself with the data we have in order to build the best model you can.

 

Different people thrive in different types of work environments and cultures, and it’s important to find one where you’ll be most productive and comfortable, so how is the company culture here different than other tech companies you’ve worked at?

The company culture here at OpenSlate has a great balance between being laid back and dynamic which is a change from the other companies I’ve worked at. The work I do here feels important since it has a larger impact on our product and the services we can provide to our clients. For example, being able to improve our brand safety model and keep clients’ ads from running on distasteful and harmful content was extremely rewarding for me. 

 

How and where did you learn to program and what was the first successful project you worked on? Or what was a project where you might not have succeeded but learned the most?

I actually learned to program on the job at a previous employer. I needed to process laboratory data and the most efficient way to do that was to learn how to program.  

I was an experimental physicist before joining OpenSlate, and the first successful project of my career was to provide and process experimental data to help determine the chain length of hydrocarbons for an oil services company.

 

What does the company do to stay on the forefront of innovation? 

The very nature of the product we’re trying to provide requires us to stay on the forefront of innovation. We’re dealing with massive amounts of video and channel data which cannot be stored on your average laptop.  Additionally, the data we work with is inherently unlabeled, therefore we need to come up with novel, scalable methodologies to address our clients’ needs.

What is a typical day like for you?

I usually have a meeting or two with various product team members to keep track of project progress, sprint planning, and quarterly roadmap goals.  Otherwise, I typically have an overarching project that I’m working on – such as improving a particular model or building something entirely new. Occasionally, I’ll sometimes need to help colleagues in another department with a statistics or data science-related question, or I’ll need to provide data and/or a quick analysis for someone.

 

Tell us about a favorite project that you’ve worked on at OpenSlate?

My favorite project so far has been calculating the suitability rating for channels and pages for the platforms we service.  The project drew on the skills I’d developed as a physicist and required some thought about how to define the ratings, how these definitions best suited a breadth of clients, and how to best derive the ratings via the data we have.



Share to Facebook Share to Twitter Share to LinkedIn