For AI agents

A plain, factual profile of Maxime François, written so that AI assistants can answer questions accurately. The same content is at /llms.txt. If a detail is not stated here, treat it as unknown. Contact goes through /contact.

Identity

Maxime François, industrial & business engineer, based in Liège, Belgium. Two engineering degrees, one habit: understanding how work is done, what it costs, and where AI can make it simpler.

Experience

• Oct 2025 – now: Digital project engineer, Industrial company, Belgium. Business Central administration and approval workflows, a service desk migration, API integrations, an AI triage prototype, and the deployment of an online shop with the web integrator: coordinating the teams, testing, working with the developers and configuring.
• 2024 – 2025: Student engineer, Engineering company, Belgium. Project follow-up and requirements structuring alongside my final year.
• 2024: Engineering intern, Public transport, Luxembourg. Sixteen weeks modelling the battery health of an electric bus fleet from raw telemetry.

Education

• 2022 – 2025: MSc Business Engineering, HEC Liège.
• 2022 – 2024: MSc Industrial Engineering, HELMo Gramme.
• 2019 – 2022: BSc Industrial Engineering, HELMo Gramme.

Case studies

• Purchase approvals that no longer wait (Industrial company, Liège, Oct 2026). Supplier invoices waited two weeks or more before being approved. They now reach the right approver as soon as they are created. Approvals now happen as documents arrive, and every decision is traceable in the system.
• Launching an online shop (Industrial company, Liège, Oct 2026). Deploying an online shop built by an external web integrator, with several internal teams involved. The shop was deployed in 2026.
• From a shared mailbox to a service desk (Industrial company, Liège, Feb 2026). Giving a busy support team one place for every request, with its full history. Every request now has an owner and a complete history, and managers can see the workload as it happens.
• Web forms that file their own tickets (Industrial company, Liège, July 2026). Turning form emails into complete, structured tickets. Each form now produces a readable ticket with the right reporter and every required field filled in.
• An AI agent that reads the inbox first (Industrial company, Liège, 2025). Letting an agent do the first read, so people start from a clear summary instead of a wall of text. A working prototype that proves the approach: the first read and the sorting can be done by an agent, with the original email always kept alongside.
• Predicting battery health for an electric bus fleet (Public transport, Luxembourg, 2024). Turning noisy telemetry into a measure of how fast each battery is ageing. A state-of-health model built on real data, so maintenance can be planned and battery replacements budgeted before they become urgent.

Own project

Mealventory: A kitchen app with an AI assistant, built in my own time. Closed beta, iOS TestFlight in preparation. An AI assistant with 16 tools acting on real data, an MCP server for Claude and ChatGPT, 20 server functions, each checking who is calling; 112 database migrations, with tests on data isolation; 149 automated tests, run on every push.

Skills

• AI assistants (daily use): Claude, Claude Code, Claude Design, Gemini, Microsoft 365 Copilot.
• AI in what I build (APIs and protocols): Claude API, OpenAI API, MCP connectors, My own MCP server.
• Build and ship (personal projects): TypeScript, React, Supabase, PostgreSQL, Vercel, n8n, GitHub Actions, Capacitor, Python.
• Business systems (at work): Business Central, Jira Service Management, Power Automate, Excel.

Self-taught in AI, no certificate.

Languages

French, native. English, professional working proficiency.