Freelancer profile translated to English.

Description

Fullstack & AI Engineer, I support my clients from design to deployment on web projects and applications integrating LLMs (chatbots, RAG agents, business automations).

My added value: the dual tech + consulting hat. I code, but I also know how to challenge a need, propose technical alternatives, and arbitrate trade-offs with you. Useful when your specs are already solid — decisive when they are not yet.

With 4+ years of experience in software development (Kitware, Thalès, AppCentrica) and several recent freelance fullstack missions, I have managed projects end-to-end: architecture, front and back development, AI integration, deployment, and production.

💡 Areas of expertise

- Fullstack web development (Vue.js + Python)

- AI Integration: RAG agents, custom tools, human-in-the-loop

- Design and architecture of tailor-made solutions

- Project scoping and technical consulting (from idea to MVP)

- Deployment and infrastructure (VPS, Docker, CI/CD)

🔧 Technical Stack

Frontend: Vue.js 3, TypeScript, Vuetify, Plotly.js

Backend: Python (FastAPI, Django), Node.js

AI / LLM: Anthropic & OpenAI APIs, RAG, Agents

Infra: Docker, Linux, Git

Technical Foundation: C++, scientific computing, 3D visualization

🎯 Do you have an AI project to launch, a web MVP to build, or a statement of work to clarify?

I will support you from the initial scoping to the production release, in a collaborative and iterative manner.

Industry field of expertise

Languages

French
Native or bilingual
English
Fluent

Workplace preferences

Can work on-site

Paris (up to 20km)

Anonyme
Production-ready RAG application on a scientific corpus
March 2026 - May 2026 (2 months)
Design and deployment of an LLM-powered augmented research web application (RAG) on a corpus of several hundred scientific documents.
Distributed ingestion pipeline (OCR parsing via LLM API, chunking, vector embeddings) orchestrated by Dagster + Celery, idempotent at partition and chunk levels (deterministic UUIDs, UPSERT).
Django/Channels webapp (WebSocket chat, hybrid vector + lexical search), multi-service deployment Postgres pgvector / Redis / Caddy via complete GitLab CI/CD pipeline (lint, tests, SHA-tagged build, SSH deploy with automatic rollback on health-check).
Python Django Dagster Docker Gitlab CI/CD
Relief
On Malt
Data extraction from STEP (3D) files
SOFTWARE PUBLISHING
February 2026 - March 2026
Paris, France
Development of an automatic analysis tool for CAD files (STEP) for industrial sheet metal assemblies. The tool extracts geometric features of parts (folds, holes, thickness, unfolded dimensions), detects contacts between parts (welded/screwed), generates a Bill of Materials (BOM), and identifies standard profiles. Exposed via a containerized REST API (FastAPI) with Docker.
Python 3D Modeling FreeCAD Docker CAD
Kitware Europe,
R&D Engineer C++/Python
September 2024 - Today (1 year and 9 months)
France
Development of open-source scientific visualization tools (VTK, ParaView) in C++/Python. Development of web applications in Python/Vuetify3 using the Trame framework. Customer support: answering questions, bug fixing.
C++ Python VTK Data visualisation 3D Rendering

Check out Jean's experience

Caroline

Directrice - SAVOIRS POUR REUSSIR GRAND EST

Several weeks project

Reviewed on 2/16/2026

Jean Fechter met our expectations, whether in understanding our needs, supporting us throughout the project, and its execution. His understanding of our needs and his responsiveness allowed us to save valuable time. The team particularly appreciated his availability and professionalism.

Mathieu

President - F3D-APP FOUNDATION

Reviewed on 11/24/2025

Another project perfectly managed by Jean despite an unexpected increase in scope and modifications to a deep part of the code that was not initially planned. We already have a project underway with Jean in succession and we are looking forward to this collaboration.

Show all 4 reviews