About Me
I am an AI/ML Engineer with a research background, currently working on applied deep learning for climate data downscaling at IPSL (Institut Pierre-Simon Laplace).
My work focuses on developing robust machine learning models for large-scale scientific datasets, with an emphasis on transfer learning, model evaluation, and physical consistency in high-resolution outputs.
With experience in medical imaging, speech signal processing, and climate modeling, I am particularly interested in building reliable and scalable AI systems that bridge research innovation and real-world impact.
Academic Background
- 🎓 Master’s in Automation and Robotics, specialization in Intelligent Systems – Sorbonne University.
- 🎓 Bachelor’s in Physics – Université Paris Cité.
Skills & Expertise
- Artificial Intelligence Research – Solid theoretical foundation combined with hands-on implementation.
- Deep Learning & Computer Vision – Design and deployment of models for scientific and medical applications.
- Medical Image Segmentation – Applied research experience at Dassault Systèmes.
- Speech Signal Processing & Machine Learning – Research experience at ISIR.
- Teamwork & Interdisciplinary Collaboration – Experience collaborating with physicians, neuropsychologists, physicists, and climate scientists in multidisciplinary research environments.
Programming Languages
- Python
- C++
- MATLAB
Libraries & Frameworks
- PyTorch
- TensorFlow
- Scikit-Learn
- NumPy
- Pandas
- Xarray
Tools & Environment
- Linux
- LaTeX
- Git
- Slurm (HPC)
Languages
- French – C2
- Tamil – C2
- English – B2/C1
Research Interests
- Physics-Informed Machine Learning – Integrating physical constraints into deep learning models for scientific and climate applications.
- Generative Models for Scientific Data – Diffusion-based and generative approaches for medical and environmental datasets.
- AI for Healthcare – Signal processing and computer vision methods for diagnostic support.
- Robust & Explainable AI – Improving reliability, robustness, and interpretability in high-stakes AI systems.
Professional Objective
I aim to contribute to high-impact AI research at the intersection of deep learning, scientific modeling, and real-world applications, particularly in climate science and healthcare. I am especially interested in environments that combine:
- Advanced AI research
- Technological innovation
- Practical solutions to major societal challenges
Publications
arXiv preprint arXiv:2604.03275
EGU General Assembly (Copernicus Meetings)
EGU26-14665
International Conference on Functional Imaging and Modeling of the Heart
Pages: 242–252
Journal of Hepatology (Elsevier), Volume 80
Page: S207
Professional Experience
AI/ML Research Engineer – IPSL
- Literature review and state-of-the-art analysis
- Use of large-scale datasets such as ERA5, CERRA, CMIP6, GeoMIP, and ARISE
- Implementation of transfer learning techniques to adapt AI models to various climate scenarios
- Evaluation of the physical plausibility of high-resolution outputs generated by AI
- Bias correction and enhancement of the accuracy of high-resolution climate projections
- Contribution to the development of a user-friendly platform to upload climate data and generate high-resolution outputs with visualization and analysis tools
Research Engineer in Medical Imaging – Dassault Systèmes
- Conducted a literature review on 2D echocardiogram segmentation without prior shape constraints
- Implemented denoising filters and segmentation methods (Morphological Snakes, UneXt, nnU-Net)
- Evaluated performance using Dice score, IoU, Hausdorff distance, along with PSNR, SNR, SSIM, and FoM for the applied filters
- Analyzed data using Nifti images
- Generated ground truth using morphological operators (erosion, dilation, smoothing)
- Aligned datasets through histogram matching
- Paper presented at FIMH 2025 (co-author): "Semantic Video Diffusion Models for Long Echocardiogram Generation"
R&D Engineer in Signal Processing and Machine Learning – ISIR (Institut des Systèmes Intelligents et de Robotique)
- Conducted a literature review
- Collaborated with physicians from La Pitié Salpêtrière Hospital (BLIPS) for speech signal analysis related to hepatic encephalopathy
- Extracted prosodic and acoustic features
- Trained models: SVM, Random Forest, Gradient Boosting, neural networks
- Developed a prediction algorithm for decision support
- Abstract presented at the EASL 2024 Congress and the 95th Scientific Days of the AFEF (co-author): "Development of a screening tool for covert hepatic encephalopathy through automated speech signal analysis in patients with chronic liver diseases and/or portosystemic shunts."
Research & Development Intern – Learning Planet Institute
- Research on birdsong mechanisms for human voice prosthesis applications
- Analysis of sound wave propagation in the bird’s vocal organ
- Study of shape–movement–sound cause-and-effect relationships
- 3D simulation and finite element modeling (FEM) using COMSOL Multiphysics
- Development of predictive aeroacoustic modeling approaches
Full CV
Photography
Photography is a personal interest through which I explore natural patterns, structure, and light. It complements my scientific perspective and curiosity about complex systems.
