Interview : Julia Renard

Optimizing our financial data system · 5 September 2024

Julia Renard, a Data Science student at EPFL, completed a two-month internship at MFM. This opportunity allowed her to explore the world of finance and apply her academic skills in a practical setting. In this interview, we delve into her experience and major achievements during the internship. As part of your Data Science master's program at EPFL, can you explain what the goal of your internship was and what your main achievements were?


Julia: Each day, MFM is receiving a large amount of financial data that needs to be treated and stored. For this, MFM has been using a data base system that has been developed gradually and through times to respond to the needs of the quantitative and investment team. This system is working perfectly but it’s not using state-of-the-art tools anymore. The idea of my internship was to start the migration and optimization of this system. I have recreated a portion of the current application using the SQLAlchemy framework to structure it into an ORM (Object Relational Mapping) model. This change in structure are changing the way we interact with the database. Instead of sending raw SQL queries to the database, we create objects that we can manipulate according to the developers' needs. This new system performs the same functions but in a different form. This is what we call “a parallel tool” to create an abstraction layer.

What are the advantages of using this new ORM model? Julia: Transforming the database using the ORM model offers several advantages. The system becomes more robust in the face of internal changes, more flexible, and easier to maintain when adding or removing features. It will also be easier to debug in the event of an unusual behaviour. Specifically, I contributed to optimizing the system by simplifying the code, improving its efficiency, speed and robustness.

How does EPFL prepare you to enter the workforce, and specifically to develop MFM's ORM application? Julia: The Data Science program at EPFL prepares us well, and we are well-equipped. The program teaches us to implement automated methods to solve problems, to consult existing documentation, and to manage new frameworks. Each situation or issue encountered is unique and requires the development of specific solutions.

Is there a significant difference between processing financial data and other types of data?
Julia: In reality, there is no significant difference. The type of data I handle may vary, but it doesn’t affect my work. I could process financial data without even realizing it. My skills are transferable to any domain and type of data.

Can you explain, your specific role in data processing ? Julia: I am not involved in data analysis. My role focuses on managing how the data is organized, processed, and transformed. Therefore, I don’t need to understand the business logic behind the data but rather the data management system itself.

During these two months, you had a glimpse of our work. What is your overall impression, and what do you take away from the world of finance?
Julia: I was impressed by the complexity and expertise required to measure the performance of various investment strategies. I realized the importance of observing performance over long periods of time to draw meaningful conclusions from all these indicators. Although it was sometimes frustrating not knowing how the data I processed would be used, I appreciated being included in the team’s weekly meetings. They helped me better understand the context in which my work is integrated. I really enjoyed this internship; it was a rewarding experience.