Freelancer profile translated to English.

Description

Data Scientist & AI Engineer specializing in the design and deployment of generative AI systems in production: LLM agents, RAG pipelines, model fine-tuning, and business NLP applications. I help you transform your use cases into concrete, robust, and measurable solutions, not prototypes that stay in a notebook.

What I can bring you concretely:

Design and deployment ofAI agentsto automate your workflows (customer support, CRM, document processing) with controlled access to your databases.
Construction ofRAG pipelineson your documents, tickets, or knowledge bases, with semantic search (Qdrant, FAISS, Pinecone).

*Fine-tuning and evaluation of LLMs(Llama, Mistral, Command R, OpenAI/Anthropic) adapted to your domain, with rigorous benchmarking.

API integration, Docker deployment, production rollout on cloud (AWS, Azure) or HPC.

What sets me apart:A dual research and production culture, fueled by international experience. A graduate of a Master's degree in Statistics, Data Science & Applied Mathematics (University of Lyon 1), I worked on the generation of medical reports by AI in Geneva (Biotech Campus), then on AI agent systems in production. French-Greek, having studied in France and worked in Switzerland, I integrate naturally into multicultural teams. I can write robust code as well as challenge technical choices based on solid statistical foundations.

Types of projects:

AI agents and business process automation
Internal chatbots and assistants (RAG on documentation, FAQs, etc.)
Generative AI POCs, feasibility validation, technical audit
Fine-tuning on specialized data
Applied NLP (classification, entity extraction, summarization)

Multilingual (FR/EN/EL/ES), I work remotely or on-site, on short missions (POC) as well as long-term support.

Industry field of expertise

Languages

French
Native or bilingual
Greek
Native or bilingual
English
Native or bilingual
Spanish
Conversational

Workplace preferences

Can work on-site

Lyon (up to 50km)

Elit-Technologies
Data Scientist / AI Engineer
TELECOMMUNICATIONS
September 2025 - Today (9 months)
Lyon, France
At Elit-Technologies, I am involved in the design and deployment of generative AI solutions in production, with the objective of automating processes with high operational load (CRM, customer support, subscription management) and accelerating decision-making.

Design and deployment of anAI agent systemautomating CRM workflows and customer support, with a significant reduction in ticket response times.
Construction of aRAG pipeline(Retrieval-Augmented Generation) with Qdrant for semantic search on internal documentation and ticket history.
Development of custom tools allowing the agent to execute SQL queries on the production database and apply advanced filters on the vector store.
Automation of ticket triage, status tracking, and subscription lifecycle management (EoS/EoL).
**Technical stack**: LangChain, LLMs, Qdrant, SQL, Python production-ready and scalable architecture.
Artificial Intelligence Agentic AI Python Docker Microsoft Azure
Campus Biotech,
Data Scientist Intern
MEDICAL
April 2024 - October 2024 (6 months)
Geneva, Switzerland
Research internship within the Data Science for Digital Health (DS4DH) group at the University of Geneva, led by Prof. Douglas Teodoro. My work focused on evaluating generative AI for automating the writing of medical reports, with the goal of reducing the administrative burden on physicians.
Design of an end-to-end NLP pipeline generating synthetic clinical reports through prompt engineering on Llama, Mistral, and Command R.
Fine-tuning and deployment of LLMs on HPC infrastructure using Apptainer and Docker.
Fine-tuning of a Longformer classifier capable of distinguishing AI-generated reports from those written by medical experts.
Rigorous evaluation of outputs with NLP metrics (F1, BERTScore, accuracy, overlap metrics).
Results contributed to a submission to EMNLP 2025.

Technical stack: Python, Hugging Face Transformers, PyTorch, Apptainer, Docker, HPC infrastructure.
LLM Fine-tuning Generative AI Docker Classification Benchmarking