Description

I help companies build AI and Machine Learning systems that drive measurable business impact.

Over the last 7+ years, I have designed and deployed large-scale ML solutions across AdTech, Real Estate and Insurance, working on revenue optimization, recommendation systems, forecasting, bidding systems, fraud detection and GenAI applications.

At Teads (ex-Outbrain), I led multiple high-impact initiatives:

Ad pacing and forecasting systems contributing to ~$12M annual revenue uplift
Deep learning optimization models used for global advertising campaigns
Large-scale ML pipelines processing billions of rows with Spark
Research and experimentation on CTR prediction, multimodal embeddings and auction optimization

I combine strong business understanding with deep technical expertise across:

Machine Learning & Deep Learning
LLMs and Generative AI
Forecasting & Optimization
Recommendation Systems
NLP & Transformers
Large Scale Distributed ML (Spark, AWS)

I am particularly effective on projects where companies need:

production-grade AI systems,
measurable ROI,
fast execution,
and strategic guidance on ML product direction.

Tech stack: Python, PyTorch, TensorFlow, Spark, AWS, SQL, FastAPI, Scala, Docker, LLMs, RAG, Transformers.

Langues

Français
Bilingue ou natif
Anglais
Bilingue ou natif

Préférences en matière de lieu de travail

Accepte de travailler sur site

Paris (jusqu’à 10 km)

Teads France
Senior data scientist
PRESSE & MÉDIAS
février 2022 - Aujourd'hui (4 ans et 4 mois)
Paris, France
• Led the improvement and scaling of large-scale deep learning models (Deep & Wide, DeepFM, Transformer-based architectures) to optimize ad performance and increase platform margins.

• Designed and conducted large-scale experimentation frameworks to evaluate and deploy advanced ML approaches, including Transformer-based models for CTR prediction and bidding optimization.

• Developed data-driven optimization features used by global advertising clients to improve campaign delivery, targeting efficiency and overall performance.

• Built bid recommendation and revenue forecasting systems leveraging time series forecasting and ML optimization techniques, contributing to significant revenue uplift and operational efficiency.
Deep Learning TensorFlow Scala Apache Spark Transformers
SeLoger Group
- Data scientist
IMMOBILIER
janvier 2020 - janvier 2022 (2 ans)
Paris, France
• Improved real estate listing localization accuracy using BERT-based Named Entity Recognition (NER), increasing conversion rates and enhancing search relevance.

• Designed and deployed recommendation systems that increased qualified leads and improved user engagement across the platform.

• Built and maintained large-scale data pipelines using Spark and Scala to process and prepare real estate datasets for ML applications.

• Developed A/B testing frameworks to measure product and model performance, enabling data-driven product optimization and experimentation.

• Created monitoring dashboards and analytics tools to track ML performance, business KPIs and operational metrics for product and business teams.
Spark Scala FastAPI Docker A/B Testing
Generali
Data Scientist
BANQUE & ASSURANCES
juillet 2018 - novembre 2019 (1 an et 4 mois)
Paris, France
• Designed and deployed fraud detection systems leveraging active learning, semi-supervised learning and anomaly detection techniques (Isolation Forest, XGBoost), helping reduce operational costs and financial risk.

• Built an OCR-based IBAN extraction system using deep learning and computer vision models to automate document processing and significantly reduce repetitive administrative workload.
XGBoost Deep Learning Computer visision Fraud Resent