We were very fortunate to host 3 talented students for a Data Science internship at MediaGamma. They'd interact directly with our engineering and data science teams to gain real-world experience working as a data scientist.
Agnes Johannsdottir (University College London, MSc. Business Analytics) was kind enough to share her experience.
Tell us a bit about yourself?
I am a 26-year old scientist and an engineer from Iceland. I love discovering new knowledge and applying it to solve challenging data-driven problems. I studied engineering at Reykjavik University and recently completed my 3 month internship at MediaGamma.
Describe the problem you're working on?
I was designing and conducting experiments on machine learning algorithms and recommender systems based on real-world e-commerce dataset in order to recommend new products based on previous behaviour. I developed an understanding of a diverse body of academic literature to motivate and ground experiments. All the development was done in Python and related open-source packages to write an extensive and stable code base. As a follow-up project, I collaborated with an e-commerce company to implement the recommender engine that I worked on during my thesis.
Describe a typical day?
Throughout this internship, the days varied. Most of the time I worked at MediaGamma offices, but every Thursday I went to the e-commerce client to make sure that all requirements were fulfilled. Furthermore, I often worked from home, usually related to writing up my thesis and research. If I got stuck with coding I had great people at the office that I could ask questions. My main work was to implement different recommender engines, do data munging, feature extraction, work on different dataset domains and figure out what model provides the best-personalised ranking for a single-and-cross domain. My supervisor was very helpful to keep me on track and focus on my research topic. He constantly challenges me and his experience is extremely valuable.
Can you describe any challenges you've had to overcome?
The biggest challenge I faced was to work with a real-world dataset. The challenges lied mainly in cold-start and sparsity problems, when trying to figure out the best model for a particular dataset.
What additional skills have you gained from this internship?
I believe my practical data science skills are wider, where the master degree is quiet academic, the change to work with a company that specialises in data science was a great opportunity. I learned to think more critical about efficiency and robustness of code implementation. Furthermore, I have worked on my python skills, where I feel more comfortable to write code now in Python and C++.
What do you plan to do after the internship? Has this internship changed your mind about your career trajectory?
I now work for AGR Dynamics in Iceland, where I will be building up a data science team in the coming months along with developing buying and merchandising software. The internship gave me great inspiration to start teaching practical data science.
Anything else you'd like to add?
The thesis gave me great insights into both academic and practical challenges in machine learning and recommender systems. I have learned a lot, where I worked with leading data science team as well as collaborate with e-commerce, which I am very grateful for.