We are thrilled to welcome Kit Rodolfa as Research Director at RegLab! Kit has spent years at the intersection of machine learning and public policy and brings a wealth of experience using novel computational tools to modernize government and to help solve some of society’s greatest challenges. He is particularly interested in work that not only increases the efficiency and effectiveness of government, but also makes it more fair and transparent through the responsible deployment of these state-of-the-art technologies. With the growing use of machine learning and data science across all aspects of our daily lives, Kit sees RegLab’s pairing of practical, impact-driven work and cutting-edge research as an ideal environment for focusing these innovations in a direction that can have the greatest benefit on society.

Kit previously worked as a Senior Research Scientist for the Data Science and Public Policy Lab at Carnegie Mellon University and University of Chicago, where he contributed to public interest-focused machine learning projects, co-developed courses for both machine learning and public policy students, and helped run the Data Science for Social Good Summer Fellowship (DSSG). He also led the initial data science efforts at Devoted Health and served as Chief Data Scientist at Hillary for America, as the Director of Digital Analytics for the White House Office of Digital Strategy during the Obama administration, and on the analytics team during President Obama’s 2012 re-election campaign. Kit holds a PhD in Stem Cell Biology and Master’s in Public Policy from Harvard University, MPhil in Chemistry from the University of Cambridge, and BS degrees in Physics and Chemistry from Harvey Mudd College.

We are grateful that Kit has decided to bring his considerable talents to RegLab and look forward to seeing the impact of his leadership, mentorship, and insight in our work.

How have you thought about charting your own path?

My path has certainly been anything but linear! I started off in college studying physics and chemistry, and though I enjoyed those subjects on an intellectual level, what felt a little lacking for me was a clear connection to the impact I might be able to have on the world. That’s really what drove my decision to change directions in grad school and spend some time exploring biology – there was a lot of hype and media coverage at the time around how revolutionary stem cells were going to be for medicine and it felt like an exciting effort to be a part of! I do believe those tools have a lot of promise over the longer term, but nevertheless felt like there was a big disconnect between the day-to-day work I was doing in the lab and those eventual real-world applications. As a bit of an escape, I started taking some classes at the public policy school and got very excited about not only the meaningful work that the faculty and students were doing but also by the way they were using data and the scientific method to design, evaluate, and improve policies and programs. It was actually a bit of a surprise to me that some of the tools I’d learned in the natural sciences could transfer to such a different domain (I remember taking an economics class at one point and thinking, “oh, this is all the same math as those physics classes from college, but they’ve renamed everything!”). So, after I finished my PhD, I decided to shift directions again and pursue a public policy degree full-time.

Of course, coming out of grad school, I now had this very cobbled-together set of skills that I imagine made my resume seem a bit confusing. Fortunately, someone on the digital team (back then, it was still called “new media”) at the Democratic National Committee thought my background seemed interesting enough to take a chance on me for an analytics internship and from there I was lucky with the timing that President Obama’s re-election campaign was just staffing up and was the first to use data and analytics in a really serious way. I learned so much there from my friends and coworkers there about different technologies and ways of thinking about and approaching problems. 

Looking back, I’m humbled by how much my experiences and the amazing people I met during the 2012 campaign opened up opportunities for me to explore different ways I could use data science to have a positive impact on society: improving access to healthcare and citizen engagement at White House in the Obama administration, leading a team of incredible data scientists on Hillary Clinton’s campaign, and working with some of the most inspiring and dedicated people I’ve ever known on two startups (one helping nonprofits make better use of social media and another trying to improve healthcare outcomes for Medicare beneficiaries). Where I really felt like my work could have the biggest impact, however, was at the intersection of academia and policy implementation, which created an opportunity for a feedback loop between meaningful, practical projects and broader research questions that were grounded in the needs of these impactful applications. I first experienced how effective work at this intersection could be with the amazing interdisciplinary team at the Center for Data Science and Public Policy at the University of Chicago (which has more recently moved to Carnegie Mellon University), and it’s a remarkable opportunity to be able to continue to work in that space with the team at RegLab.

I guess all of that is to say that in a lot of ways, I don’t think of my career path as having been particularly charted or well-planned, so much as the result of a lot of exploration and learning along the way. I’ve always wanted to pursue work that felt like it was meaningful, impactful, and challenging, but it took a little while to figure out exactly what that meant to me. It may have been a circuitous path, but I’m so glad it’s led me to RegLab and am excited to be a part of the incredible work that’s happening here!

What are your hopes for what RegLab can achieve?

I think RegLab really occupies a pretty distinctive role that doesn’t exist at too many other places. Sitting at the intersection of research and practice creates some real opportunities to scale the impact the group can have beyond what would be possible, for instance, if it existed entirely within a given government agency or focused on more traditional (and more abstract) research. 

There’s a really productive feedback loop that can develop between the lab’s applied, project-focused work and its broader research directions. Not only does our work with government agencies and non-profit organizations provide a direct route to social impact through their implementation, but it also tends to very quickly highlight gaps in how well existing methods and academic knowledge apply to real-world settings. These gaps, in turn, generate new research questions and hypotheses that can help us refine these tools in order to improve their utility in application. Particularly in fast-moving academic areas like data science and machine learning, there’s no shortage of potentially intellectually-fascinating research questions that one could pursue, but working from the context of the on-the-ground needs of an organization that’s trying to have a real impact can really help focus your research in a useful direction!

However, building those partnerships is hard and takes a considerable amount of time and effort to develop the necessary trust and relationships with the people at these organizations. I think RegLab’s problem-centered approach has been incredibly important in this regard – rather than focusing on a particular method or tool and seeking out users to try it out, we focus more on the core challenges faced by our project partners and seek out the right tools under the broader umbrella of data science that fit those needs. At the same, being housed within Stanford gives us access to a wide range of amazing technical and scientific talent that facilitates this flexible approach. 

Taken together, all of this makes RegLab a pretty special place for being able to do impactful work you can feel good about while at the same time addressing some of the core questions at the forefront of scientific research. I think that really feeds into how I think about the four broader goals I see the group as especially well-positioned to accomplish:

• Large-scale positive social impact, both directly through the organizations we work with, but also more broadly by demonstrating how this work can generalize across a range of policy domains and contexts.
• Training a new generation of scientists and technologists on the interesting challenges and nuance of working in applied, socially-impactful settings and inspiring them to continue to work in this space.
• Educating our partners (and the broader policy community) about how their data can be put to use to help meet their goals, how to be better consumers of technical tools, and the value in improving their own data and technical infrastructure.
• Contributing to scientific knowledge, with a particular focus on research questions driven by what is most practically useful and helps us develop and deploy the tools of data science in a responsible, socially impactful manner.