|
||||
|
||||
|
|
For this edition's spotlight we interview Rob Chew, Class of 2014. Rob came to the IAA from Boston and has made his home in Raleigh, NC where he has joined a salsa team and caught quite a few pokemon on the RTI campus. Ashley: What group do you work in at RTI International? Rob: I work in the Center for Data Science (CDS), a relatively new organization within RTI International. Clay Heaton and I were the first data scientists hired into the group two years ago. Since then, CDS has expanded to 20 people, six from the IAA! (Peter Baumgartner '14, Rob Chew '14, Clay Heaton '14, Lauren Klein-Warren '12, Mahin Manley '15, Paul Ruddle '14). Our team consists of data scientists, software/web developers, data visualization specialists and statisticians, many whom also have subject matter expertise in an applied natural or social science discipline (crop science, economics, public health, etc). Ashley: Wow, I didn't realize there were so many IAA alums working at RTI! Rob: We're no SAS or M&T Bank, but we've got our fair share :) To my knowledge, two other IAA alumni also work at RTI -- one in the Sustainable Business Solutions practice (Christina Beauchamp '13) and another in RTI Health Solutions (Theresa Coles '08). Not to mention, there are probably close to 10 alumni and/or current students who worked at RTI before starting at the IAA (Kelsey Campbell '16, Viola Glenn '16, and Chris Liedtke '16, to name a few). Ashley: So you're a Program Manager? Rob: Yes, I split my time between a technical role and supervising employees; I manage a group of six data scientists and web developers. Ashley: What variety of projects do you work on? Rob: Our project work is very diverse, both in subject domain and types of data analyzed. For example, this morning I had a kick-off meeting for a project where I'll be creating interactive data visualizations with hospital data tax data. For lunch, I met with a collaborator to discuss a remote sensing pilot project that uses images from drones to identify breeding grounds for mosquitoes, in prevention of Zika virus. In the afternoon, I juggled between modifying one of my Python scripts that collects tweets relating to e-cigarettes and writing a proposal blurb for NLP applications regarding EHR clinical notes. Other folks in CDS are doing awesome things on projects relating to criminal justice, transportation, international development, education... it's really all over the place. Ashley: How does the Center for Data Science fit within RTI? Rob: RTI has a long and storied history of data analysis and statistics. It was founded as a non-for-profit research institute by the triangle universities of NCSU, Duke, and UNC Chapel-Hill and was the first occupant of the Research Triangle Park (RTP). As an illustrative example, Gertrude Cox, the founding head of the NC State Statistics Department and the university's first female full professor, was also the first head of RTI's Statistics Research Division. For better or worse, this means that not many folks at RTI are impressed by our knowledge of statistics -- we've had to make a name for ourselves by taking on non-traditional analytics problems with text, images, and streaming data and combining it with web development, interactive visualizations, and human-centered design. When CDS was created, we were given a mandate to integrate and promote "Data Science" across RTI, meaning we work with whoever will work with us and has data: criminologists, engineers, neglected tropical disease specialists -- you name it. We partner with other groups within RTI that have their own external clients and who want to add more to what they're currently offering. The other groups bring the subject matter expertise to provide the context and important questions within their domain. By engaging them and the client, CDS in essence plays an analytic consulting role, finding where we can add value with tech, data, and design. In addition to the functional groups, we've also been successful in engaging with RTI management and administrative groups like HR on internal business intelligence work. Ashley: What technologies do you use? Rob: We primarily use open source software, with the exception of Tableau and occasionally SAS. We're pretty much a Python shop when it comes to data work but it really depends on the problem and, to an extent, what our collaborators are comfortable with. R is also great if we're doing one-off analyses using domain specific methodologies that don't yet have a Python library. For instance, we're creating an R package on a new propensity score matching algorithm that a group of RTI health economists developed. We use some flavor of the common web technologies (HTML, CSS, and Javascript) for visualizations & web apps, though I'd be lying if I said I had in-depth knowledge of all the web languages and frameworks used on our more development-heavy projects. Though most of the data scientists can throw together a Flask app and know some d3.js, we leave most of the heavy development work to our software engineers. Those who do more data visualization may use languages germane to computer graphics and creative coding such as WebGL, Openframeworks, or Processing, in addition to d3.js. We also experiment with new data stores or "Big Data" processing technologies and try to incorporate them into projects when it makes sense (MongoDB, Neo4j, Spark, etc.). Ashley: So you just came back from Indonesia with RTI - what did you do there? Rob: I was there for 16 days of initial engagement / vacation in May, where Clay and I met various stakeholders and started with scoping the project. We were there to create a pilot educational platform that will help districts analyze and visualize the data they send to the national government, to make programmatic decisions. This trip to Indonesia is part of a larger UN & World Bank initiative of "Data Revolution for Development". The idea is to get better and more accessible data for the developing world to aid in policy making and to use technology to leapfrog development efforts. We started in Indonesia because RTI has been there for 25 years and we have over 500 employees in the region. Indonesia is in a good place timing-wise, since there's support from the national and local government and a growing user base for technologies like smart phones. Believe it or not, Jakarta holds the title of the world's most active tweeting city! 
Taking a break during with monkeys and orangutans during a free weekend in North Sumatra, Indonesia. |
|
|
