Farouk E.

A training organization needed a complete system to attract, capture, and convert prospects — without daily manual intervention.

I designed and deployed a complete digital infrastructure from A to Z:

→ High-conversion landing page — Next.js & Tailwind CSS, optimized registration forms, and custom user journey

→ Headless WordPress + GraphQL Blog — articles fetched via GraphQL and displayed on the Next.js landing page to maximize SEO without content duplication

→ AI Content Pipeline — automated article generation via ChatGPT (OpenAI), visual creation with Cloudflare AI, continuous automated publishing

→ Complete automation with n8n — lead collection and processing from forms, intelligent routing, automated follow-up emails, and nurturing

→ Integrated e-commerce store — ebook sales with Stripe, state management with Zustand, smooth and secure purchasing experience

→ Industrialized production deployment — Docker containerization, automated CI/CD pipeline for zero-downtime updates, hosting on OVHCloud

Result: A complete system in production, generating organic traffic, capturing leads, nurturing prospects, and selling — 24/7, without manual intervention.

Design and deployment of an AI assistant capable of instantly answering team questions based on their internal documents — in natural language, without manual searching through files.

How it works:

Documents are stored on Azure Blob Storage, processed, chunked, and vectorized in Azure AI Search (semantic search), then queried via an LLM deployed on Azure AI Foundry — all exposed via a FastAPI API and a modern Next.js interface.

What I built from A to Z:

→ Complete RAG pipeline: document ingestion from Azure Blob Storage, chunking, vectorization, and semantic indexing in Azure AI Search

→ LLM integration via Azure AI Foundry: the model receives the extracted context and generates a precise, sourced answer

→ Secure REST API with FastAPI (Python) to orchestrate the entire AI pipeline

→ Next.js Frontend — intuitive chat interface for end-users

→ Authentication and user management with C#/.NET + Supabase database (PostgreSQL)

→ Complete cloud deployment and access security on Azure

Result: A production-ready RAG solution, multi-user, secure, and scalable.

SOCOMEC, a major industrial player, needed to digitize and secure the reception of its visitors across 3 industrial sites with full GDPR compliance and zero manual intervention.

I designed and deployed a complete end-to-end solution:

→ Two Power Apps applications: an Agent interface for administrators (visit management, Excel import, history, export) and a multilingual FR/EN touch kiosk for visitors (self-registration, GDPR consent, badge printing)

→ 6 automated Power Automate flows: badge printing via secure REST API (Bearer Token), Teams and email notifications, automatic GDPR anonymization, archiving, and PDF generation

→ C#/.NET REST Webservice deployed on IIS: instant printing of visitor badges via Brother QL printer, secured with HTTPS + Bearer Token

→ Multi-site architecture: 3 independent environments (U1, CIL, U3) with centralized administration and automatic routing based on the connected site

→ GitLab CI/CD Pipeline: automated build, controlled deployment across 3 environments (development, staging, production), rollback possible

→ Complete Security: Microsoft Entra ID, RBAC, MFA, HTTPS, logging of all actions

Result: 12 check-ins in 60 seconds via kiosk, GDPR compliance validated, zero manual intervention on reception processes.