Hi Arne, could you introduce yourself? Tell me where you are from, what your educational background is, if you had any prior work experience before your internship at HCSS?
My name is Arne Eichholtz, I am Dutch, born and raised in Maastricht. My background is in AI, I did my Bachelor in Artificial Intelligence at the Radboud University in Nijmegen, from which I graduated last summer. As for previous work experience, I did a summer internship during my studies but other than that little professional work experience. HCSS is my first proper 9-5 job.
By the end of June, I will have been at HCSS for almost 9 months as an intern in the data lab. The regular internship is six months, but I was given the opportunity to extend, so I gladly decided to extend until the summer.
So nice that you were able to extend your internship. I’m curious what initially attracted you to the HCSS internship programme, and why did you apply specifically for the Datalab?
About a year before applying, I came across HCSS on Twitter, and I had seen the name before. Looking into what HCSS did, I saw that it’s a thinktank with projects about defense and foreign affairs, mostly for the government. Interestingly, there was a page about the HCSS datalab, it listed an old internship vacancy. I decided to either contact HCSS myself or apply to a new vacancy when the moment was there. Luckily, when I was looking for an internship a year later, there was a regular vacancy and I applied and received the position!
HCSS fits my interests since it combines qualitative and quantitative research. During my studies I followed some Political Science courses, I love reading about politics and generally I am quite politically interested. Thus, I am drawn to combine my technical background in policy settings. HCSS was the perfect fit.
Can you describe your time as a Datalab intern? What does your day-to-day look like?
I was hired for the professionalization of the Natural Language Processing (NLP) pipeline. The goal is/was to build a generic data pipeline that works on a variety of different text data sources, for instance reports, debates, or news articles that can complement statistics on, say, crime. We are working on a pipeline because the idea is that it takes text corpora as input, and the output is a visual or dataset. This means that we build all the steps in the process: data analysis but also data preprocessing and building out visualization tools.
Some work had been done on the NLP pipeline, but not a lot. And since NLP was a new research area for HCSS, I was given the opportunity and freedom to shape the development of the project. If I came across a new method, tool or visualization, I could just pursue it after consulting with my supervisors if it fitted reasonably within the project or goal.
Some of the work includes building software to extract text from PDFs or building an extensive topic modeling analysis.
In the beginning the day-to-day was more focused on generally building out the pipeline: what techniques are being used? How do we integrate it into our work? What do we use it for? In the past several months, however, there were several projects for which the pipeline was to be used. All projects require tailored input, analysis and output, and the day-to-day has thus focused on the needs of the project application and client. Questions like: what data is required? How should it be cleaned? What should the output look like? It is pretty cool: the stuff I built now has been used in several projects for clients.
What did you enjoy the most about your internship?
Taking initiative is appreciated and cherished at HCSS. If you have a lot of ideas, there is a lot of freedom to explore within the parameters of the project and this has been great.
Additionally, it was inspiring to be part of the smart, talented and dedicated team at HCSS. Moreover, there are a lot of interns and since you’re all new together, there is a sense of camaraderie. We go out for drinks almost every week, and this is good fun. Besides the interns, there are a lot of young people at HCSS and there is a good atmosphere at the office.
How did the team or your supervisor/mentor contribute to your internship experience?
The relation with my supervisors has been great. As I have explained, it was up to me to shape the development of the project, but always in discussion with my supervisors. The communication has been superb, and general atmosphere as well.
What are your next steps after completing this internship?
Some months ago, I decided that I will shift away from the application of AI in policy settings for now, and instead double down on the technical AI aspects, namely with a technical Master. I will start the AI Master at the University of Amsterdam in September. And I have plans to go to the US for grad school after that.
This interview was conducted by HCSS Communications & Events Assistant, Stephanie Govaerts
Experts
Related Content
