Resume

Safouane Chergui

Lead Data Scientist | Paris, France

chsafouane@gmail.com | LinkedIn | GitHub

Technical Expertise

ML/AI Domains: NLP/LLMs, RAG Systems, Search & Recommendation Engines, Computer Vision Cloud & Infrastructure: GCP (Vertex AI, BigQuery), AWS, Docker, Kubernetes, MLflow, Airflow Specialized: LLM Evaluation, Production ML Monitoring, Model Drift Detection, Agentic Systems

Professional Experience

Founder & AI Engineer | LumiereAI | April 2025 - Current

  • Built production agentic system for insurance companies using Google ADK framework and Vertex AI Gemini API
  • Designed multi-modal pipeline integrating ElevenLabs voice synthesis with Gemini-powered insurer case analysis
  • Architected end-to-end solution on GCP: Cloud Run for serving, BigQuery for analytics, Cloud Storage for document ingestion

Lead Data Scientist | EDF | April 2025 - Current

  • Architected document parsing and RAG pipeline on GCP using Vertex AI embeddings and Vector Search for semantic retrieval over internal knowledge bases
  • Built LLM evaluation framework leveraging Vertex AI Model Evaluation to benchmark Gemini and custom models on business-critical tasks
  • Deployed production inference endpoints using Vertex AI Prediction with automated scaling and monitoring

Senior Data Scientist | Doctolib | Sep 2024 - April 2025

  • Architected and deployed a RAG system reducing customer support costs by 20%
  • Enhanced phone assistant system that allows patients to make a medical appointment using a fast search system and agents

Senior Data Scientist | Mirakl | Oct 2023 - Aug 2024

  • Built & deployed an LLM-powered catalog integration system reducing products onboarding time from 20+ days to hours
  • Implemented end-to-end MLOps pipeline for model versioning and automated retraining using MLflow
  • Trained & Deployed small LLMs for domain-specific tasks to reduce costs of third-party API usage by 10X

Senior Data Scientist | TheFork (TripAdvisor) | May 2022 - Oct 2023

  • Designed recommendation algorithms driving 35% increase in page visits across web, mobile, and email channels
  • Pioneered LLM-based review summarization system processing 500K+ reviews monthly
  • Optimized search functionality using semantic embeddings (Elasticsearch, pgvector)
  • Built collaborative filtering and content-based recommendation models serving 20M+ users

Data Scientist | EDF | Dec 2018 - May 2022

  • Deployed production ML models for customer segmentation, churn prediction (AUC 0.92), and LTV modeling
  • Designed real-time drift detection system monitoring 15+ models in production environment
  • Built A/B testing framework and interactive dashboard reducing experiment analysis time by 60%
  • Developed predictive maintenance models using signal processing and deep learning (CNN/RNN architectures)

Key Achievements & Leadership

  • Production Impact: Deployed 20+ ML models to production including Vertex AI endpoints with measurable business and costs impact
  • Technical Leadership: Mentored 5+ junior data scientists, conducted Python/R training sessions at EDF
  • Cross-functional Collaboration: Led projects spanning engineering, product, and business teams

Education & Certifications

  • MSc Engineering - INSA Lyon (2013-2018)
  • Deep Learning Specialization - Coursera
  • Statistical Learning - Stanford Online

Languages

  • English (Native)
  • French (Professional)
  • Arabic (Native)
  • Spanish (Fluent)
  • Mandarin (In the learning process)