CV

Experience

• Research Intern, Shakhnovich Lab, Harvard University, April 2024 - August 2024
  • Run well-tempered 100ns metadynamics for protein complex binding free energy prediction, using GROMACS and PLUMED on a Slurm HPC environment
  • Evaluate energy minimization MD methods to predict binding free energy using ESM3 as encoder for SVM regressor: “Does Structural Information Improve ESM3 for Protein Binding Affinity Prediction?” Accepted at Machine Learning in Structural Biology Workshop at NeurIPS. I demonstrated an unexpected loss of around 0.2 Spearman Correlation using relaxed structures compared to wildtype structure.
• AI Software Engineer Intern, Pretto, June 2023 - August 2023
  • Managed deployment and access control on platforms such as GCP, Azure AI, and Heroku.
  • Focused on optimizing user adoption and ease of use across multiple features, including:
    • Automated Document Generation: Created automatic summary reports from client files using GPT.
    • Call Transcription and Summarization: Leveraged Whisper and GPT to transcribe and summarize phone calls.
    • Bank Statement Analysis: Analyzed bank statements using Python libraries (pandas) and text extraction from PDFs.
    • Internal Knowledge Search Engine: Researched and implemented Retrieval-Augmented Generation (RAG) techniques to enable a search engine that retrieves and generates answers from company knowledge, using diverse content sources. Implemented with LangChain and Pinecone.
    • Interactive Slackbot Development: Built an interactive Slackbot to improve internal workflows.

Projects

• RNA Clustering with Graph Neural Networks, Bioinformatics Project, 2024 Work done with Yann Ponty, Sebastian Will and Johannes Lutzeyer. I use a GNN encoder (Residual Gated Graph Convolutional Network) to cluster RNA secondary structures. The model achieves a RSME of 3 on the distance prediction, which correponds to a actual error of 3 base pairs on a total of 50, considered as an acceptable error for clustering.

• Fake News Detection, Deep Learning Project, 2023 Work done with Ekimetrics. In a team of 4, we train a Ensemble Model using Machine Learning algorithms (Linear Regression, XGBoost…), LSTM model with Word2Vec and BERT model to detect fake news on the dataset ‘LIAR’. Bert achievs a 70% accuracy on the test set.

• Segmentation of lung cancer cells in 3D CT Scans, Institut Polytechnique de Paris Medical Data (IPPMeD), 2023 Work done in a team of 2. We build a 3D U-Net model using Monai framework to segment lung cancer. We obtain a F1 score (Dice Metric) of 0.5 on our test set and 0.37 on the hackathon test set. We achieve the 2nd best score on the hackathon.

• Reinforcement Learning for the CartPole problem, 2023 I implement a Q-Table and a Deep Q-Network (DQN) to solve the CartPole problem. The Q-Table achieves a score of 250 on average, while the DQN achieves a score of 100 at most, facing convergence issues.

• Silo Detection in Satellite Images, Ecole polytechnique x QuantumBlack Hackathon, 2022 Work done in a team of 4. We developed a ResNet network to detect silos in satellite images (93\% of accuracy), delivering a real-time, globally-operable prototype within 24 hours and using Grad-CAM for segmentation. We won the hackathon by achieving the best score on the test set.

Education

• Applied mathematics, Ecole polytechnique, 2021-2025
  • Honors: Outstanding Investment and Outstanding Leadership awards (given to around 10% of the most active students).

  Courses mainly in applied maths and computer science

  • Foundations of Machine Learning
  • Stochastic Models in Finance
  • Operational Research: Mathematical Aspects and B Applications
  • Image Analysis and Computer Vision
  • Advanced Deep Learning
  • Random and statistical modeling of processes
  • Random Models in the Environment and Evolution

  Additional coursework in the health/biology domain:

  • Drug Sciences
  • Computational analysis of high-throughtput sequencing data
  • Bioinformatics Project in clustering of RNA secondary structures using GNN encoders
• Deep Learning, MVA (Mathematiques, Vision, Apprentissage), ENS Paris-Saclay, 2024-2025 Courses in Machinea Learning:
  • Reinforcement Learning
  • Topological Data Analysis
  • Time Series analysis
  • Computational Statistics
  • Convex Optimization

Skills

• Programming Languages: Python, C++, OCaml, Java, SQL
• Programming Frameworks: PyTorch, Scikit-Learn, Monai, HuggingFace, spaCy, LangChain, Pandas, Numpy
• Technologies: Deep Learning, Computer Vision, Natural Language Processing, Git, GCP, Slurm

Extra-Curricular Activities

• Student Union representative in charge around 150 clubs, managing event planning, legal matters, and accounting, while being responsible for an annual budget of 2 million euros
• As the Artificial Intelligence Treasurer, I played a key role in organizing the annual hackathon in collaboration with QuantumBlack at Ecole polytechnique on detection of methane detection