Description

Tech lead en machine learning et NLP, je suis votre guide vers l'avenir numérique. Avec un doctorat en Large Language Models et une expérience diversifiée couvrant la finance, la santé, le juridique et le marketing, je transforme vos données en puissantes insights/modèles et agents. En tant que votre partenaire stratégique, je suis prêt à propulser votre entreprise vers de nouveaux sommets grâce à l'IA.

Languages

French
Native or bilingual
English
Fluent

Workplace preferences

Remote only

Primarily works remotely

Waalaxy
Gen AI Tech Lead
July 2022 - Today (3 years and 11 months)
Montpellier, France
The Leads Generator: A recommender system leveraging RAG architecture with optimized vector search across 700+ million LinkedIn profiles and 40+ million companies, with enhanced contextual filtering.
Waami: An advanced LLM-based multi-agent system generating highly personalized prospecting sequences (notes, messages, emails with 3 follow-ups) with improved context management between touchpoints.
The Lead Scoring: Dual high-performance classification models predicting both new conversion potential and expansion opportunities with existing clients, featuring confidence scoring and regular retraining cycles.
Interest Sentiment Analyzer: A refined model categorizing prospect responses into interested, not interested, or potentially interested later, with improved boundary case detection.
Fine-tuning several 7B LLM models for prospection messages, with domain-specific optimization and evaluation metrics tailored to the effectiveness of sales messaging.
All of these systems are deployed in production through robust pipelines handling preprocessing, inference, training, and comprehensive observability metrics with automated performance monitoring.
Agent IA LangSmith Langchain Prompt engineering Cloud GCP Docker Grafana Fine-tuning LLM
Université de Bourgogne
Ph.D Student
December 2019 - June 2022 (2 years and 6 months)
21000 Dijon, France
Topic : Automatic writing of abstracts from texts using lexico-grammatical patterns and deep learning
• NLP/NLU : Keyword and domain-based Lexicogrammar patterns (LxGR) extraction, Text classification, Corpora tagging, Extractive and Abstractive text summarization, Data augmentation.
• Studied domain : Finance. Models: BERT, UniLM, BART, CamemBERT, GPT-2, GPT-J, T5. Libraries: Pytorch, Tensorflow, Keras, NLTK.
• Python
• French/English
Université de Bourgogne
Lecturer Machine Learning
September 2020 - March 2022 (1 year and 6 months)
21000 Dijon, France