Freelancer profile translated to English.

Back to original language

Description

🧠 Generative AI Expert — Agents, RAG, Synthetic Data, Fine-Tuning & Automation

I assist you in designing, integrating, and optimizing generative AI solutions tailored to your business needs.

🔹 What I do

1. Creation of advanced AI agents

Autonomous or orchestrated agents

Complex conversational workflows

Integration of APIs, databases, business tools

Multi-agent / agentic systems architectures

2. Design of RAG systems (Retrieval-Augmented Generation)

Building complete pipelines: ingestion, chunking, embeddings, vector store

Optimization of the RAG graph (system graph)

Performance evaluation (RAG triage, hallucination minimization)

Integration into web applications, chatbots, or internal assistants

3. Synthetic data generation

Creation of realistic datasets for training / QA

Improvement of data quality or diversity

Anonymization & GDPR compliance

4. Model training and Fine-Tuning

LLM fine-tuning (LoRA, QLoRA, SFT, DPO)

Training of specialized models (classification, extraction, NER, summarization…)

Cost/performance optimization (quantization, pruning)

Deployment on servers, cloud, or local

🔹 For what needs?

Intelligent internal assistant

Automation of business tasks

AI-augmented search engine

RAG chatbot on internal documents

Data analysis and processing

Custom AI model adapted to your sector

🔹 Why work together?

Hands-on experience on several AI projects in production

Technical quality + pedagogy

Fast, clear, and documented delivery

Robust, performant, and scalable solutions

Languages

French
Native or bilingual
English
Fluent

Workplace preferences

Remote only

Primarily works remotely

noet-ia
Research and Development Engineer
December 2024 - Today (1 year and 6 months)
Grenoble, France
Main responsibilities

Design and development of autonomous AI agents
Creation, orchestration, and optimization of agents capable of performing complex tasks, integrating decision-making, planning, and interaction mechanisms with APIs or internal systems.

Implementation of RAG systems (Graph RAG / System Graph RAG)

Building knowledge graphs to improve information retrieval.

Integration of advanced RAG pipelines to combine databases, embeddings, search engines, and LLMs.

Optimization of retriever performance, indexing, and source structuring.

Creation of synthetic data

Generation of synthetic datasets for model training and evaluation.

Implementation of data augmentation and enrichment strategies.

Verification of the quality, consistency, and diversity of produced data.

Training and finetuning of AI models

Supervised training, instruction tuning, RLHF, or DPO depending on project needs.

Optimization of models for specific tasks (NLP, classification, generation, agents).

Monitoring metrics, managing datasets, and continuous improvement of ML pipelines.
LIG - Laboratoire d'Informatique de Grenoble
CIFRE PhD Student in Computer Science
September 2025 - Today (9 months)
Grenoble, France
Develop specialized NER models

Adaptation, finetuning, and evaluation of LLMs on complex entity extraction.

Design of annotated datasets (real and synthetic).

Model and extract relations (RE)

Implementation of models capable of identifying and classifying relations between entities.

Development of multi-task NER + RE strategies.

Explore automatic generation and ontology enrichment

Automatic formalization of concepts, properties, and relations.

Use of LLMs to propose, validate, and update ontologies.

Align NER/RE results with a knowledge graph

Integration into Graph RAG or KG-enhanced LLM architectures.

Detection of inconsistencies, disambiguation, and symbolic reasoning.

Implement industrial pipelines

Optimization, deployment, and evaluation of models in a production environment.

Automation of workflows via LLM/AI agents.