Work History
Advanced Financial Market Analysis Research Project
March 2025
Researcher
Developed an innovative unsupervised approach for anomaly detection in financial time series using two complementary deep learning architectures: a custom temporal convolutional autoencoder (TCN-AE) and a modified transformer model. Leveraging 50 years of market data (1973-2023) from CRSP covering 36,328 different assetsImplemented and compared three distinct training strategies to optimize model detection capabilities
Discovered that models trained on bearish market windows demonstrated superior discriminative capacity, effectively identifying market configurations preceding upward movements (1.61% average returns for transformer model, 0.76% for TCN-AE)Designed and back-tested a simple trading algorithm based on detected anomalies, achieving average returns of 4.40% (transformer) and 3.07% (TCN-AE)Demonstrated temporal persistence in signal generation, suggesting the models captured fundamental market structures rather than statistical noiseThis research revealed interesting asymmetric behavior in market predictability and offers promising perspectives for algorithmic trading strategies.
TransformersAuto-Encoders
Loox
Started in 2022
Entrepreneur - Founder - Developer
I initiated the creation of Loox, a social network dedicated to fashion. This entrepreneurial project allowed me to showcase my versatility as a full-stack developer, taking charge of the complete development of the application.Using React Native for the mobile application, combined with Java for the backend, I crafted a robust and efficient system. To ensure data integrity, availability, and speed, I opted for Cloudflare R2 for storage, Redis for caching, and PostgreSQL for the database.I implemented a sophisticated recommendation algorithm based on user interactions, leveraging embedding technologies to analyze user preferences and behavior patterns. This feature significantly enhance the user experience by providing personalized content suggestions, increasing engagement and time spent on the platform.Website
Safran Aircraft Engines
Feb 2024 - Aug 2024
Research Data Scientist Intern
During my internship at Safran Aircraft Engines, a global leader in aircraft engine manufacturing, I was tasked with implementing an advanced AI for engine anomaly detection. This involved the descriptive analysis of complex flight data, encompassing millions of flights from worldwide partners.I enhanced the existing predictive model accuracy by 4%, building upon a project initiated by senior data scientists. Collaborating closely with two senior specialists, we explored new research directions that significantly deepened our understanding of the model outcomes and improved overall performance.My role included refining algorithms by leveraging cutting-edge technologies such as AutoEncoders and Transformers. This approach pushed the boundaries of what we could achieve in aircraft engine diagnostics, contributing to Safran's position at the forefront of aviation safety technology.
AutoEncoderTransformers
Prediction of Bridge Creep
2023
Researcher - Developer
During this research project, I had the opportunity to dive into the challenge of predicting bridge creep using a particularly limited data set, a situation that required adopting specific strategies and models. In collaboration with a civil engineering expert, we sought to maximize the utilization of this valuable data. Recognizing the limitations of traditional approaches due to data scarcity, we opted for a hybrid method.By pairing physical equations, which provide a fundamental understanding of bridge behavior, with machine learning models, we were able to produce more accurate and reliable predictions. This fusion of physical principles and artificial intelligence not only improved the accuracy of our predictions but also opened up new perspectives in approaching civil engineering problems with limited data.
Fullstack Engineer
2019 - Present
Developer Freelance
For several years, I have showcased my passion and expertise as a freelance developer specializing in web and mobile domains. My skills span a variety of technologies, including React, React Native, Java, Redis, PostgreSQL, Solidity, and more. I have also had the opportunity to work on intricate projects in the decentralized finance (DeFi) space, allowing me to refine my technical approach and familiarize myself with the unique challenges of this booming industry.My freelance journey has enabled me to assist a diverse range of clients, from innovative startups to more established companies, providing them with tailor-made solutions that cater to their needs while incorporating the best practices of modern development. Each collaboration is an opportunity for me to learn, innovate, and contribute to the success of the projects I engage in.
Education
University Of California
2023
Studying Data Science for three monthsat the University of California in San Diego (UCSD)where I attained a GPA of 3.9/4.UCSD is renowned for its academic excellenceand is notably ranked 19th globally in the Shanghai Ranking.
Gustave Eiffel University
2021 - 2024
Graduate of ESIEE Paris engineering schoolholding a French "Diplôme d'Ingénieur"(Engineering degree, recognized internationallyas equivalent to a Master's degree with adistinctive focus on applied engineering)
| Subject | Rank | Total |
|---|---|---|
| AI and Deep Learning | 3 | 20 |
| Algorithmic | 14 | 136 |
| Cybersecurity | 8 | 135 |
| Advanced Network | 1 | 11 |
| Java & C | 13 | 78 |
| English | 25 | 135 |
Preparatory school
2019 – 2021
A two-year post-A level/secondarylevel course in math and physics inpreparation for the competitiveentrance exams for M. Sc Courses atengineering universities