In honor of Women’s History Month, I had the great pleasure to interview Ysis Tarter. She is not a familiar household name—yet!

Her path to data science began from a desire to help people. Initially drawn to the medical field, Ysis was accepted as an intern with the BioSTEP program at Yale’s School of Medicine.

“There was a scientist in the lab working on something, and it was so tedious,” she recalls. “I remembered from a class I had taken in the spring that if you just modify a couple lines of code, you could save a lot of time. I did it for him and realized you can really speed up and make science more efficient. That is how I ended up getting more into computer science.”

And she hasn’t looked back. Ysis is a young professional who continues to defy the insurmountable odds and works hard to pave the way forward for the next generation of young girls who will succeed her.

During my interview with this rising data superstar, Ysis said, “There is a lack of data capture in the black community. It can’t be optimized if it is not captured. If the data is there, the lack of analysis restricts the data. People are starting to tell the story, but we can’t change the story until someone starts writing the story.”

Her profound statement penetrated my soul, giving me confirmation that there is more work to be done within the Black community. And let’s begin with writing her story so we can raise, inspire and see more Black females like Ysis making a difference.

There is a lack of data capture in the black community. It can’t be optimized if it is not captured.

Why did you choose data analytics?

I started in medicine. Medicine brought me into data science. What’s important is that the population’s health is tightly linked to both.

I realized that to be a doctor is to affect on a one-to-one scale, but if you can do research and synthesize all the data being collected, you can really affect populations of people and be really effective.

Another reason was there were not a lot of people who looked like me in the field. Part of my frustration with medicine was in how things weren’t being evenly distributed. I can find the cure all I want, but are the people whom I want to get it getting it? And how would I even know it? That’s when data science comes into play.

You’ve been working in data science for a while. Run me through your background.

There are some very interesting things involving data. One of the first things I was using it for was personalization and recommendation systems—being able to synthesize a ton of people’s data into useful information.

I was a research and advanced engineering intern in product development at Ford Motors. I used IOT data science to do proof of concepts. For example, if you had on a wearable like a Jawbone or Fitbit for a day and overnight, when you got into a car I could predict your driving performance and how well you would do on an absent scale and on a relative scale, and predict on how you would rate yourself. Then you can learn cool properties of things from big data. I learned a lot of little things, like the moment right before everyone makes a left turn their heart rate speeds up. I think it is because you’re going into traffic. A lot of cool trends came out of data science when I was at Ford. This was all before they launched the bikes.

I was previously a software developer at Athenahealth. Using data science, I made recommendations to create an app-based personalization system specifically for doctors. They used it as a reference to make sure they did not prescribe drugs that might have adverse interactions.

I then moved over to the bio field, which I have always liked. As a software engineer at Zymergen, we created a landscape of different properties that could be synthesized and designed molecules that would exhibit those properties based on the data that already existed. In dealings with the LIMS (Laboratory Information Management System) and a design-build-test-learn cycle, there is data science at every point—from design to being able to render results to answering questions like, “What was previously successful? What does that look like?” We have to get a lot of components quickly: To get through large data and analyze and process that quickly is an imperative task.

We use data science to connect to a bunch of things. IOT is very popular. I also noticed that there was not enough being done, being recorded.

You’re currently working toward a Master of Science in Applied Biomedical Engineering at The Johns Hopkins University. As you continue with your own studies, what new trends in data science are you excited about?

I am excited about data science in medical diagnostics. Not only is data science finally breaking through and helping diagnose people, but it is also correlating different genetic expressions to symptoms of diseases and figuring out what is going on.

I am really excited about what we are doing in medicine. One of the tools I’m excited about and what we are using now is blockchain. When it comes to blockchain in data science, it handles data integrity and dealing with all of the nuances needed for medical record handling.

Blockchain is great for that: An immutable record that can’t be changed later, and that can be at your level of privacy discretion.

The saying, “You don’t know if you truly know anything until you can teach it” forced me to test my knowledge.

You started teaching for us in May 2019. What draws you to teaching?

Extension reached out with the opportunity for me. It was great for me because it was part time.

Another thing was personal development, staying abreast of the most current tools and also testing my own knowledge of data science. The saying, “You don’t know if you truly know anything until you can teach it” forced me to test my knowledge.

I also enjoy helping to generate responsible coders. Coming out of these programs, sometimes applying techniques and theory in the industry doesn’t quite translate perfectly. So being able to help with that disconnect is good. 

What is it like teaching adult learners?

Adults were the only people I had not taught yet! I had experience working with elementary and high school students. I taught elementary kids with Black Girls CODE and high school students at SMASH, a UC Berkeley math and science honors academy held in the summer.

It is very important to me to share the resources that I am involved in because adult learners make a more immediate impact on the actual workforce. It’s nice to see the results quickly and how you can affect those results directly.

Also, adult learners bring really cool experiences to the classroom. I always tell my students, “You're an expert at something!” You can specialize in something, you can be the best in what you bring your experiences. There was one student who had an accounting background and she went into data science financial management. I had another student who liked teaching math, and he ended up leveraging his skills for the school district.

I try to encourage students toward what they are interested in and what they are good at.

I can walk through code in real time; if bugs come up, that’s actually a good thing because they get to see how I handle it and now they can do it, too.

How would you describe your teaching style?

My teaching is a super live demo. I really like to go through things. I do a lot of live coding so that my students can ask questions. I can walk through code in real time; if bugs come up, that’s actually a good thing because they get to see how I handle it and now they can do it, too.

Visualization is good for my students. The visualization is being able to draw a conclusion from the data and share it. 

There are not a lot of people who look like me in this field. It’s not too late to educate, and at the same time it’s never too early.

What are some of your main takeaways you’d like your students to have when they complete your course?

Data science is a heavy tool—it is powerful and should not be taken lightly. With adult learners, I hope that I am teaching people who are going to go and see what the world is. What are they looking at? Are they looking at the right things? What are they leveraging?

I also want them to be aware of skewed data and echo chambers. You have to be aware of where your data is coming from. If it is not being recorded, it is not happening. You can’t fix something that isn’t happening.

There are not a lot of people who look like me in this field. It’s not too late to educate, and at the same time it’s never too early. Experience is experience.