About
Hello! I'm a master's student in computational neuroscience at EPFL. I'm currently a visiting student at MIT in the Edelman Lab fitting transcriptomics. My main research interest is deep learning applied to biology and recently, I am also interested in a more formal understanding of data which can be anything from information to random matrix theory.
This website was originally intended to archive my projects but will progressively include personal items such as quotes and jokes.
Journey
18.05.2025
I finished my internship at Metadvice and am very grateful to André Jaun and his team. Metadvice's pipeline elegantly mixes many statistical methods from decision trees to reinforcement learning to produce robust and hallucination-free neural networks. I developed a solution for comorbid networks which led to an EASD conference submission for novel therapeutic approaches for patients with Type 2 Diabetes and Chronic Kidney Disease. On top wonderful lunch time discussions about data sparsity or morality, this experience allowed me to gage the magnitude of developing software with probabilistic models.
In february, I published my first paper which was accepted at MIDL 2025 for the work of Sequence models for continuous cell cycle prediction. Recently, I received my first citation and I am beyond proud of this milestone. This work treats cell cycle prediction as a video classification task, we show this improves the performance of existing methods two fold while modeling a richer continuous representation of the cell cycle.
At the Van De Ville Lab in Geneva, I worked with Ilaria Ricchi on spinal cord graph learning using Laplacian mixture models. After weeks of frustration and poor results, debugging with synthetic data revealed limitations in Gaussian mixture models - findings I hope to develop into a future paper after strengthening the mathematical foundation.
In a few weeks, I will be moving to Boston for 8 months to start my master's thesis at MIT in the Edelman Lab. Supervised by Farhan Khodaee the project aims to use large language models to analyze single cell RNA seq data. Working on genomic data and from it predicting various phenotypes has been my goal since I dropped out of computer science at 18. Eager to learn from the ambitious minds at MIT, I hope to pursue at PhD in applied deep learning or apply for full time position at Metadvice. I am also accepting my interests have outgrown glorified curve fitting and am curious to how I approach these nexts projects.
09.09.2024
Following a rich learning experience at the Naef Lab, I developped a regression model that integrates sequential cellular brightfield images and predicts the cell cycle phase from the FUCCI system. This led to a paid summer internship, where we explored the limits of this model. As I resume my education, I will stay in close contact with the lab and this project, with the goal of a publication.
This fall, I will be starting an exciting 6-month internship at Metadvice. A start-up where I will be tasked to implement new clinical decision support models. The main focus will be comorbid situations and investigating model explainability with Shapley values. I am glad to be returning to tabular machine learning and hope to contribute some ideas like KAN-interpretability. I will also be starting a shorter semester project with Ilaria Ricchi, benchmarking novel implementations of graph mixture models, where I hope to learn more graph theory and connectome analysis.
I am glad to be applying my skills in a fitting field and hope to continue in search of a master's thesis for the spring of 2025.
06.02.2024
I'm a 2nd year master's student at EPFL studying Neuro-engineering and I'm currently looking for an internship to finish my studies.
In the spring of 2024, I will be joining the Naef Lab of computational biology as a master's student to investigate latent space models for cell cycle trajectories with live imaging data. My goal will be to implement self-supervised learning techniques and explore the utility of "nested" vision transformers to make use of the 3rd dimension.
This website catalogues my neuroscience, data analysis and machine learning projects. My objective is to use my computer science skills to aid the medical engineering revolution, leveraging deep learning tools to solve large quantitative biological tasks.
