MLOps — Lending Club
End-to-end MLOps pipeline for peer-to-peer loan default prediction
Python Scikit-learn LightGBM XGBoost Optuna MLflow FastAPI Airflow Docker Streamlit
// About
AI Engineer finishing a Master's in AI & Big Data at ESGI (work-study), with 3 years of experience at La Poste in data science and engineering. Focused on agentic systems and LLM integration in production — LLM/SLM benchmarking on Databricks, MITRE ATT&CK-based detection pipelines with an LLM SOC analyst. 2nd place at the Ministry of Armed Forces AI Hackathon (1st detection rate).
Experience
Data Scientist (Work-Study)
Groupe La Poste · Data & AI Division
Sept 2024 – Present · Paris
- › SLM vs LLM benchmark on Databricks for a job offer classification use case
- › Customer clustering and impact analysis on large-scale business data
Data Analyst (Work-Study)
Groupe La Poste · DOT IDF Colissimo
Sept 2023 – Aug 2024 · Paris
- › Automation of reporting pipelines (manual Excel → Power BI)
Education
Master's in AI & Big Data
ESGI
2024 – 2026
Work-study
Bachelor's in AI & Big Data
ESGI
2023 – 2024
Top of specialization
DUT in Computer Science
IUT Amiens
2020 – 2022
// Projects
Clair Obscur
AI-augmented NDR platform for cybersecurity anomaly detection and resolution
Python LLMs FastAPI AWS React ClickHouse PostgreSQL
Deep Reinforcement Learning Library
Rust library for comparing reinforcement learning and deep reinforcement learning algorithms on toy environments
Rust Python Macroquad Jupyter
// Skills
Languages
Python Rust SQL TypeScript Bash
AI / ML
LLMs RAG Agents Scikit-learn HuggingFace PyTorch LightGBM XGBoost Optuna MLflow OpenAI / Anthropic API
MLOps & Infra
Docker Kubernetes FastAPI AWS Databricks Airflow Terraform Git Linux
Data & Viz
pandas Dataiku DSS Spark polars NumPy PostgreSQL ClickHouse Power BI Tableau Streamlit
// Contact
Interested in working together or have a question? Reach out directly.