Author Spotlights

Author Spotlight

In the Author Spotlight series, TDS Editors chat with members of our community about their career path in data science, their writing, and their sources of inspiration. Today, we’re thrilled to present our conversation with Julia Nikulski.

Julia is currently in the final semester of her master’s degree in sustainability management and works as a research assistant at a sustainability think tank in Germany. After completing a bachelor’s degree in corporate management and economics, she worked as an analyst in the financial industry with a focus on sustainable finance. In 2019, she completed the nanodegree in data science by Udacity.

---

Thanks for chatting with us, Julia! As a grad student and researcher, what are the topics you’re most drawn to these days?

I am particularly interested in the areas of sustainability, finance, and various topics related to data science. In the sustainability field, the circular economy, sustainable consumption patterns, and the intersection of sustainability, digital transformation, and data science are issues I especially care about. On the data side, I enjoy working on NLP problems, time series forecasting, and data visualization. One of my latest projects implements the Longformer model to analyze biographical data. And an area I am currently learning more about is unit testing in machine learning.

Taking a step back, how did you find your way into data science?

While working as a financial analyst, I had virtually no programming skills. I had done some university projects in R and had started a Python course on Codecademy, but never found the time to develop actual knowledge and skills. However, I noticed how vital at least a general understanding of a programming language would be to do my job more effectively. My employer also encouraged their workforce to gain technical skills and an increasing number of data scientists were hired.

At the end of 2017, I was determined to give programming another shot. Simultaneously, I read up on what a data scientist actually does. While I started with HarvardX’s CS50 course to gain a basic level of computer science understanding, I was intrigued by the data scientist role and wanted to learn more. This coincided with my wish to take some time off and consider a career change. I quit my job and started learning data science full time.

What was the most difficult part about taking the leap into a new field?

The most challenging period of time was actually starting. Time was always a limiting factor for me. Learning a new skill is time-consuming and doing it with a full-time job can be very challenging. I was in the privileged position to be able to quit my job and learn full time. However, this was not an easy decision. I was met with quite a bit of skepticism and criticism when I told people I quit a secure job to learn skills on my own for which other people gain a university degree. I started to doubt myself and at the same time felt pressured to learn these skills fast to demonstrate that I had made the right decision.

What did you do to move past that challenging period?

A number of things helped me to overcome these issues. First, I frequently reminded myself that it is never a bad idea to invest in education. Even if I did not get hired as a data scientist, I would gain new knowledge and skills that would benefit me in any future job. Second, I created a learning structure that ensured I knew what my long-term goal was while reaching smaller, more easily achievable goals along the way. For me personally, structure is very important and I need to see the progress I am making in order to stay motivated and work effectively.

Third, I read about people’s learning journeys and career changes into data science on Medium and TDS. While stories about going from zero coding skills to being hired as a data scientist in six months did not really calm my nerves or take away the pressure, it helped me to read about other people’s experiences and struggles when starting out in the field.

Finally, I had a life outside of learning. I know what works for me and what doesn’t. Having excessive work or learning schedules for an extended period of time that prevent me from relaxing and doing things I enjoy is counterproductive. While I felt pressured to progress fast, I intentionally scheduled time off and breaks to keep a balance and ensure I could stick to my goals.

What advice would you give to people who are just starting out their data science journey?

Think about why you want to learn about data science. Understanding what you want to get out of this learning experience can help you set goals, stay motivated, and achieve an outcome you are happy with. Goals should be ambitious, yet achievable. It can be demotivating if you have one very ambitious goal which you cannot meet.

Adjust your learning to fit your needs. As I mentioned, I need to have structure and I learn best with short video tutorials and frequently applying gained knowledge in small projects. Knowing how you learn most effectively will make acquiring new skills easier. And finally, I would recommend being patient. While some people are able to become a data scientist in six months, putting too much pressure on yourself can lead to frustration or even giving up on your goals.

What inspired you to write for a broader audience?

I was introduced to writing for Medium and TDS during my nanodegree at Udacity where I had to write two blog posts about my projects. I kept writing beyond that because I enjoyed the process of explaining concepts and creating tutorials to help others and to expand my own knowledge. Moreover, writing technical articles about machine learning models, for example, can demonstrate that I know what I am talking about. This is especially important for me, given that I am pursuing data science outside of a university degree. Being able to point to something and say, "I can apply this and understand how it works," even if I didn’t learn it as part of a university degree, is very valuable.

Do you have any insights to share with people who might be considering writing for a larger audience?

If you want your story to resonate with an audience you need to be informed and able to communicate potentially difficult concepts. I notice with every article I write that I expand my own knowledge. When I write about machine learning, in particular, I always read academic journal articles to understand the implementation and reasoning behind using a model. I truly enjoy this process of gaining and communicating knowledge. I am also happy about the connections I’m making with people who read my stories, find them useful, or have comments and reach out to me.

One last question: what changes would you like to see in the data science field in the next couple of years?

I hope that more people in the data science community think about the positive impact their skills could have on people and on the planet. The public sector, NGOs, and communities can benefit from skilled data scientists. I also hope that the topics of data ethics and data privacy will become a more integral part of the data science field and education. I personally want to learn more about these areas to become a more responsible data scientist.

Finally, I would like to see that the data science community continues to grow because I believe that this field benefits from a plurality of perspectives and backgrounds.

---

If you’d like to explore more of Julia’s work and wide-ranging interests and projects, visit her Medium and GitHub profiles. Here are some of her publishing highlights from our archives.

• Data Science for Sustainability (TDS, January 2021) This post shows readers how data science and sustainability work can intersect and inform each other, and includes several compelling real-world use cases.

• 5 Steps to Develop Unique Data Science Project Ideas (TDS, September 2020) If you’re experiencing an inspiration slump, Julia’s practical suggestions for kickstarting a new project can help. They run from generating new topics based on your daily routines to finding data sets that haven’t been extensively used by others.

• Time series forecasting with AdaBoost, random forests and XGBoost (TDS, February 2020) What do you do when your model doesn’t produce neat, perfect results? In this post, Julia shares the useful lessons she’s learned while working on a particularly challenging project.

• How to Build a Data Science Portfolio Website (TDS, May 2020) With a competitive market to navigate, job searchers and career switchers need to highlight all the relevant skills and experience they have. Here, Julia walks us through the process of building an effective data science portfolio site, and covers both setup and content ideas.

---

Stay tuned for our next featured author, coming soon! (If you have suggestions for people you’d like to see in this space, drop us a note.)

Share This Article

• Share on Facebook
• Share on LinkedIn
• Share on X