Freelancer profile translated to English.

Description

I am interested in missions in data science, artificial intelligence, or prompt engineering. I have significant experience on several projects involving retrieval-augmented generation (RAG), language models (LLM) like Mistral or Llama, speech to text, and recommendation algorithms.

I am open to freelance opportunities.

Furthermore, I publish a newsletter that simplifies artificial intelligence and has over 3200 readers.

Languages

French
Native or bilingual
English
Fluent

Workplace preferences

Can work on-site

Bruxelles (up to 50km)

Orra Academy
AI Engineer
March 2023 - Today (3 years and 3 months)
Paris, France
I am CTO of Orra Academy. I define and manage the technical strategy, design product architecture, and ensure their security, scalability, and quality. I supervise developers, implement processes, and integrate best practices. I continuously explore new AI models and protocols (LLM, RAG, agents, automations), testing and then integrating them into our offerings. I also support client projects to ensure the success of technical deliverables.
LLM Generative AI AI Agents RAG Artificial Intelligence
Amaris
AI engineer /LLM / Data
PRESS AND MEDIA
October 2022 - January 2025 (2 years and 3 months)
Bruxelles, Belgium
Project 1: Media Content Prediction and Recommendation

Objective:
Create a machine learning algorithm to predict the lifespan of newspaper articles and develop recommendation algorithms for video podcasts, articles, and other media content.

Tasks:
• Develop and implement a retrieval-augmented generation (RAG) algorithm to enhance recommendations and content retrieval processes.
• Build and optimize recommendation algorithms for video podcasts, articles, and other media types.
• Utilize large language models like Mixtral-8x7B for text processing and understanding.
• Integrate Whisper large v3 for transcribing audio and video content, enabling better indexing and search.
• Train a CamemBERT model for article classification and relevance prediction, using sentiment analysis to enrich metadata.
• Annotate and preprocess data for various media types, design scripts to increase dataset size for more robust model training.

Tech Stack: Python, Whisper, Mixtral-8x7B, AWS, Kubernetes, LangChain, Docker, Neo4J & Terraform

Project 2: R&D on LLMs

Objective:
Study LLM capabilities, conduct a comparative analysis, and explore ChatGPT applications.

Responsibilities:
• Automate podcast transcription via the ChatGPT API.
• Implement intelligent podcast clipping using GPT-4 and Spacy.
• Create titles for different media platforms.
• Develop targeted descriptions for the audience.
• Work on image generation with Dall-E and optimize images using ChatGPT for prompts.
• Enhance social media post creation for platforms like Facebook, LinkedIn, and Instagram with ChatGPT.
• Train consultants on the ChatGPT API.
Python LLM Artificial Intelligence AI Agents Langchain
Marketing 1by1
Junior Data Engineer
August 2021 - August 2022 (1 year)
Lille, France
Objective: Optimize and standardize the club's data model for current and future football sector clients. Main Responsibilities: ETL Process: Capture structured and unstructured data. Integrate data using Python, MongoDB, and JavaScript. Manage data mapping and quality (standardization, deduplication, anonymization). Collaborate with stakeholders to validate data. API and Datamart: Establish data mappings. Database Development: Design database segmentation tools. Documentation: Standardize data integration procedures. Monitoring: Use tools like CloudWatch to track data evolution, alerts, and statistics. AWS Usage: Explore and implement tools and services within the AWS ecosystem.