Freelancer profile translated to English.

Description

Hello,

Looking to automate a time-consuming process, integrate AI into your product, or quickly develop a full-stack feature that's production-ready? That's exactly what I do.

I am a full-stack developer and n8n expert based in Toulouse. I work with SMEs, startups, and freelancers who want to save time with automation, integrate LLMs (Claude, OpenAI, Ollama) into their workflow, or develop a production-ready Next.js + FastAPI product.

What sets me apart: I built DevisMatic (devismatic.fr) alone, a B2B SaaS in production for French construction professionals. It transforms a construction site photo into a PDF quote in 2 minutes using AI. In practical terms, I know how to deliver a product that stands up to real clients, not just a pretty prototype that breaks on first use. I'm aware of the pitfalls: authentication, GDPR, error handling, French compliance, and deployment on real infrastructure.

Types of projects I work on:

→ n8n Automation: Connecting tools, eliminating manual tasks, workflows triggered by forms, webhooks, or APIs. 1 to 3 days.

→ AI / LLM Integration: Data extraction from documents, structured generation, business chatbots, automatic classification. Anthropic (Claude), OpenAI, or local Ollama APIs. 2 to 5 days.

→ Full-stack Development: MVPs, new features, module redesigns. Next.js + FastAPI + Supabase stack, deployed on Vercel, Railway, or Docker. 3 to 10 days.

How I work: Complete autonomy, I frame the project with a few initial questions, I deliver in small verifiable steps, and I frankly state what's feasible and what's not.

I remain open to any proposal, even outside these three areas if the subject is interesting. Feel free to contact me to discuss it, I respond quickly.

Available immediately, in Toulouse or fully remote anywhere in France. English B2.

See you soon,

Pierre Jouin

Industry field of expertise

Languages

French
Native or bilingual
German
Basic
English
Conversational

Workplace preferences

Can work on-site

Toulouse (up to 50km)

DevisMatic
Founder & Fullstack Developer
CIVIL ENGINEERING
December 2025 - Today (6 months)
Toulouse, France
DevisMatic (devismatic.fr)

B2B SaaS for French construction professionals: plumbers, electricians, masons, painters, plasterers, roofers, carpenters, and all other construction trades.

The problem: A small construction business owner loses an average of 5 to 10 hours per week writing quotes after work. Time they don't bill for, which delays proposal delivery, and often leads to losing deals to more responsive competitors.

The solution: DevisMatic transforms a construction site photo into a professional PDF quote in about 2 minutes. The tradesperson takes a photo from their phone, the AI analyzes the scene, identifies the work to be done, suggests quote lines with market prices and quantities, and generates a ready-to-send PDF.

Beyond generation, the product manages the entire sales cycle:
→ Integrated Assistant for adjusting quotes in natural language
→ Client CRM with complete history
→ Factur-X compliant invoicing (mandatory French standard)
→ RelanceAI: Automatic reminder module with penalty calculation according to French commercial law
→ Customizable quote and invoice templates (Classic, Modern, Compact)
→ Secure and GDPR compliant storage

Technical Stack: Next.js 14 (frontend), FastAPI Python (backend), Supabase (database + auth + storage), Stripe (payments), Resend (transactional emails), Anthropic API (Claude for the AI engine), self-hosted n8n on Hetzner for automation workflows. Deployed on Vercel + Railway. OWASP/ANSSI/GDPR security audit passed.

Business Model: Freemium (3 free quotes per month) with a 7-day Pro plan trial, then a monthly subscription.

What I learned: Building a product from scratch that addresses a real need, delivering with complete autonomy, managing French legal compliance, and integrating AI properly into a demanding business workflow.

Launched in 2025, in the phase of acquiring first paying customers.
Artificial Intelligence Next.js FastAPI Business Strategy Administrative Management
DSI,
AI Developer Intern
HUMAN RESOURCES
January 2026 - February 2026 (1 month)
Toulouse, France
Development internship at Distribution de Services Industriels (DSI) in Toulouse, with complete autonomy under the supervision of Sky Gunning.

Objective: Design and deliver an internal tool for automatic CV anonymization and standardization, to facilitate HR processing by removing personal information while preserving the professional structure and useful content.

Delivered solution: A complete Python pipeline that ingests CVs in various formats (PDF, DOCX), extracts text, identifies personal entities using a local LLM, anonymizes sensitive data (name, contact, address, photo), and then restructures the content into a standardized format usable by recruitment teams.

Technical Stack: FastAPI for the API, Ollama with llama3.1:8b for the local LLM (zero external data transfer, GDPR constraint), SQLite for persistence, Docker Compose for orchestration, all deployed under WSL2.
Delivered Volume: Approximately 4,700 lines of code across 19 files, complete technical documentation, unit tests on critical functions.

What I learned: Architecting an AI application autonomously from specifications to deliverable, deploying an LLM locally to ensure confidentiality, managing GDPR constraints on sensitive data, and delivering within a short timeframe (6 weeks) without daily supervision.

Mission validated by the internship supervisor and the technical team.
FastAPI API Integration Artificial Intelligence Docker
I-Cube Research (Groupe I-Pulse)
R&D Intern
MECHANICAL ENGINEERING
April 2025 - July 2025 (3 months)
Toulouse, France
Research and development internship at I-Cube Research, part of the I-Pulse Group, on the I-Terra project: applying artificial intelligence to precision agriculture.

Objective: Develop an automated plant detection module from field images to help farmers identify cultivation areas, measure plant cover, and detect anomalies without manual intervention.

Work performed: Image processing pipeline in Python with OpenCV, segmentation using HSV color space, filtering and morphological operations to isolate plant areas. Complete Dockerization of the pipeline with serverless deployment via Nuclio under WSL2 to enable on-demand execution on new images.

Integration with CVAT (Computer Vision Annotation Tool) for annotating training and validation datasets. Automation scripts for generating JSON annotation files, managing imports/exports, and testing the model in the field.

Deliverables: Operational pipeline, complete technical documentation, end-of-project presentation, and engineering-level validated thesis.
Stack: Python, OpenCV, NumPy, Docker, Nuclio, CVAT, JSON Schema, WSL2.

What I learned: Translating a field problem into a technical solution, working in an R&D environment, handling computer vision in production, and delivering documentation usable by other developers or researchers.
Docker Artificial Intelligence API Integration Business Strategy Database