

Discover more from Health Data Science Newsletter
A Q&A With Director of Data Science Kelly Burdine at Wellthy
If you are Health Data Scientist Manager or Director, let me know if you are interested in sharing the story of your career journey. You can email me at ahobby@healthdatasciencenewsletter.com
Summary
Andrea Hobby interviewed Kelly Burdine, the Director of Data Science at Wellthy. Wellthy is a company that provides personalized support and advocacy services to individuals and families managing chronic, complex, or aging-related health issues.
Can you talk about your career journey thus far to become a director of data science? How'd you get there?
Until recently, there were no data science programs or anything. It's not like I went to the data science degree and got a job in data science. These weird routes are interesting, but I was in the Air Force. I was on active duty Air Force years ago. I worked in intelligence reconnaissance, so I did a bit of analysis there and spent a lot of time collecting large amounts of data. And when I got out of the Air Force, I worked as an analyst for the government as a government contractor, so I started to dig deeper into the analysis. And at the time, I was working in a master's program, and they had a track for data science. That's when I started to get exposed to coding more quantitative stuff and different methods there. Then, I got asked to be part of this research group focused on applying different data science methods to solving problems in international relations. And so I started working for them part-time, and most of it is like self-taught, a lot of the like, technical skills and things that I have, and I loved it. I thought it was great. I decided I didn't want to work in government anymore. I ended up going and getting a job in tech. I've been a tech ever since, and I love working in tech. So I've worked in sports technology in tech, and now I'm in the health tech space, and I've had the good fortune of being the first data hire at a couple of different tech startups. So I've learned a lot on the job a lot along the way of how to, you know, how to build processes and infrastructure and things that scale well because it's one thing to build a machine learning model and analyze something, here's just one problem, but how do you do it like repetitively in a way that you can manage and maintain. Anyways, in the first tech company I worked at, I was the first data hire and grew to a team of 15. So obviously, at some point, I took on a leadership role. Then I've been in leadership roles ever since. And so I've been at Wellthy for a little over a year. They had the start of a data function and brought me in with my experience managing and growing data teams.
What qualities do you look for when hiring for your data team? How do you evaluate them during the hiring process?
I'm looking for SQL skills from a technical standpoint. Another common experience and some other coding languages are usually good. I don't generally care what it is because if you can learn one, you can learn them all. Data visualization skills are another important technical skill. Also, it depends on the role. Some roles are more engineering focused, where you need modeling experience. Some roles are more analytics focus where you know; you might need more statistics. So that depends a little on the soft skill—the ability to communicate and tell a story with data. Curiosity is a big one. And then attention to detail. So those are the soft skills on my second floor.
What challenges have you faced with growing your data teams? How do you address these challenges?
One is, well, technology is always changing. Right? I'm always looking to be able to adopt new technologies and take advantage of them. So, it's more of a fun challenge. I mentioned this before, but building something that scales well. Something might work for a smaller team, company, or project. How do you build it when talking about petabytes of data coming through? How do you know how it will perform when it's running every single minute in production? So, there are these challenges of scalability that suddenly you need different methods for doing. Then there's always a challenge around like there's always more data questions that are asked that you can't answer, like there's your data team always feels bigger, right? Like you never have enough bandwidth to handle all the data needs. And so, prioritize and move fast, but not in a way that creates a ton of tech debt. And so it's about, you know, smartly, prioritizing and thinking about things long term, versus, you know, is this just a one-off thing that I need to do or is this something I'm going to have to do again and again?
How do you stay up to date with the latest trends and technologies? How do you ensure that your team also stays up to date?
I follow a couple of newsletters and roundups that consolidate some of the latest blogs and findings. So Data Elixir and The Analytics Engineering Roundup is another one. There are tons of them. Then I'll find certain people I like the stuff they put out, and I'll follow them. The biggest way I stay updated with changes and trends and what's going on is through data communities. There are a couple of really good data Slack communities out there that I've learned a ton from and gotten a ton of help from. And the like, one is called Locally Optimistic. That's my favorite. And then there's another one. It's a dbt Slack community, so the dbt hosted by the dbt labs team goes far beyond that. The topics go far beyond that specific tool, but there are a number of them out there. That is a great way to stay up and understand how other people use the different tools and technology they're creating and everything.
As far as keeping my team up to date, we use Slack internally for communication and have a data trends channel. I encourage people to post there. So people drop stuff in there all the time and say, "You know, hey, here's a really good thing I listened to or watched or read," and "I think we could use this or, you know, I like I don't agree with this one part but, this piece over here we can use." It is important, especially as the team leader taking the time to do that and encourage that sets the right culture for the team that learning is important. We don't want to go with the status quo, and it's worth spending a little bit of time each week learning new things and pushing that, so that's one thing we do. We also use Guru. It's like a Notion or Confluence internally. We have a tools and technology section. So when we find an interesting new tool or something, we drop it into that. So we've created this repository of interesting things that we found.
Anything else you'd like to share for early career data scientists who might be interested in becoming a data science director one day?
One thing this is true for people who have started their careers or are trying to break into the field is to help people solve a problem. Right? So I hear this a lot where someone's like, "Yeah, you know, working in accounting, and I did this boot camp, or I'm taking some classes online, and I want to get in there like how do I do that? Because every position out there needs experience, I don't have any experience." And what's going on at your company? What are some problems you see that you can help solve, right? So find opportunities in your current company, even if it's not part of your role, where you can leverage data, get involved, and solve a problem. It's a great way to start getting your hands dirty and getting that experience. Then, you can do personal projects outside of work if you still need to, but experience is important. And if you help people solve problems and you make an impact, they're just going to keep coming back to you. They'll see you're going to become valuable.
Links:
https://locallyoptimistic.com/community/
https://www.getdbt.com/community/join-the-community/
https://dataelixir.com/