Freelancer profile translated to English.
Back to original languageAbout Tristan
Expert Data, Machine Learning & Generative AI
Data Scientist and AI Engineer, I assist companies that want to leverage their data.
My experience as a Data Scientist allows me to intervene from research to production, combining technical performance with business impact.
⭐️ My key expertise
- **Generative AI**: Agentic AI - multi-agent, MCP, RAG (retrieval augmented generation), LLM fine-tuning (pytorch). Expertise in Langchain, Langgraph, and CrewAI
-Machine Learning & Data Science:fine-tuning, optimization, and deployment of predictive algorithms in NLP and Computer Vision, python
-MLOps, Data Engineering & Industrialization:pipelines (dbt, Dataform), CI/CD (GitLab CI), containerization (Docker), IaC - infrastructure as code (Terraform), Backend (FastAPI)
-Cloud & Data Platforms:GCP, AWS, Snowflake, Dataiku.
🤖 Types of missions
-Tech leading:Architecture, best practices, team coaching, code reviews, knowledge transfer.
-Development of AI products up to production:AI & multi-agent Agents (LangGraph, CrewAI, MCP), RAG, LLM fine-tuning (vision, NLP), prompt engineering.
-Data & Backend Engineering:data pipelines (dbt, Dataform), APIs (FastAPI), CI/CD, infrastructure as code (Terraform).
-Multi-cloud context:GCP, AWS, Snowflake, Dataiku — from POC to industrialization.
📁 Recent projects (dev)
-AI Agents + MCPto automate CRM campaign creation at Accor
-Industrialized RAGto leverage hotel support documentation databases
-Multimodal LLM Fine-tuning(vision-text) for photo scoring
-Hybrid Cloud LLM (GCP/AWS)for image SEO → €100K saved, +€1M in revenue
-Outfit Recommendation:end-to-end industrialization for LVMH/Celine
I help you transform your data into high-performing, scalable, and maintainable AI products.
French
Native or bilingual
English
Fluent
Spanish
Basic
Can work on-site
Paris (up to 50km), Grenoble (up to 30km), Lille (up to 10km), Rennes (up to 10km), Toulouse (up to 10km)
Experience
- StealthLead AI EngineerSOFTWARE PUBLISHINGSeptember 2025 - Today (9 months)Paris, France
Context
Lead AI Engineer for an early-stage startup in a team of 4.Achievements
- Development and production deployment of generative AI & multi-agent systems
- Development of an agentic architecture optimized for text writing (reducing response time by a factor of 4)
- Context Engineering - drastic reduction in the number of tokens processed (125K -> 25K tokens for one of the agents) and improvement of workflow robustness
- Development of LLM output validation middlewares
- Implementation of almost all AI & backend software infrastructure
- Scoping and requirements specification
- Code reviews
- Implementation of tests (unit and integration) - including agentic tests
- Advanced prompt engineering through collection and analysis of business technical documents
Tech stack
GCP, docker, python Langchain, Langgraph, CrewAI, FastAPI, Convex, FastAPI, Kubernetes, pytest, asyncio - ACCOR SALead Machine Learning EngineerHOSPITALITYApril 2024 - September 2025 (1 year and 5 months)Issy-les-Moulineaux, France
Context
Long-term assignment as Lead Data Scientist / Machine Learning Engineer within the Accor group's Data Science team.Example Achievements
MCP: Automation of CRM campaign creation processes-> Estimated Savings: ~€40K/year in direct savings + campaign creation time reduced from several weeks to a few hours- Reverse engineering of the SalesForce Marketing Cloud (SFMC) Content Builder data model, a CRM platform for campaign creation
- Open-sourcing of a python client for SFMC
- Implementation of utilities to analyze and manipulate the SFMC Content Builder data model
- Implementation of an MCP server allowing campaign creation from a compatible AI client (Claude Code, ...)
Tech stack: python, asyncio, pydantic, mcp python-sdkLLM: Improvement of photo scoring model-> Decision support greatly improved by a performance increase of 30% to 130% (depending on metrics) compared to the previous model- Complete redesign of a photo annotation platform (+10K photos annotated from 30 different brands)
- Technical management of a junior data scientist responsible for fine-tuning a new model
- State-of-the-art VLM fine-tuning specialized in aesthetic scoring. Total dataset of 300K images
- Evaluation and quantization of the fine-tuned model
- Production deployment (300K images scored/month)
Tech stack: AWS, pytorch, streamlit, transformersChatbot / RAG: Generative AI at scale-> Helping numerous Accor teams benefit from AI-augmented products, through specialized assistants possibly augmented by RAG.- Implementation of RAG (Retrieval Augmented Generation) for various business contexts (Hotel Support, ServiceNow, Design team...)
- Creation of specialized AI assistants for business teams (Communication, Legal, Procurement, etc.)
Tech stack
Python, AWS, GCP, Docker, Pytorch, LangChain, Terraform, Streamlit, dbt, SQL - OCTO TechnologyData ConsultantDIGITAL AND ITJanuary 2022 - March 2025 (3 years and 2 months)Paris, FranceA data consultant role (data scientist / data engineer / ml engineer / cloud).Numerous missions carried out in varied business contexts (Public Finance, Hospitality, Luxury, Insurance...) as well as complementary roles held (Data Engineering, Data Science, MLOps, Tech Leading).Tech stackAWS, Snowflake, GCP, python, docker, dbt, terraform
Recommendations
NB
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- MSc Autonomous Vehicle Dynamics and ControlCranfield University2019MSc Dynamique et contrôle des véhicules autonomes
- Mines EngineerMines Nancy School2019Spécialisation information et systèmes
Certifications
- Professional Data EngineerGoogle2024