Description

Most "automate this for us" conversations start with the wrong question. The real one is: which process is actually worth automating, and which isn't?

I run a fixed-price, 2-3 day AI Process Diagnostic: map the process AS-IS, find the one costing the most time or money, and hand back a TO-BE flow with a clear yes/no on whether AI is the right tool for it. No discovery-phase invoice creep.

Agentic RAG tool (LangGraph + Claude + ChromaDB, multilingual) that runs full EU AI Act compliance audits end-to-end: risk classification, GPAI detection, gap assessment, client-ready PDF report, fully local for data privacy. Also shipped UrbanScore (Le Wagon): image-based property search + price prediction with LLMs and HuggingFace embeddings.

When a Process Diagnostic surfaces a regulated AI system, I can run that same compliance gap-check as an add-on — not a separate practice, just the natural next question once you've mapped how a process actually works.

Why this combination holds up:

→ European Commission (ATOS/DIGIT), 2.5 years: designed the RBAC and approval-workflow architecture for an EU-scale ServiceNow migration — who can request, what triggers automatically, what escalates. The governance question agentic AI raises today, solved at EU scale before the term existed.

→ BNP Paribas: GDPR data governance on high-security banking platforms — data mapping, risk classification, stakeholder alignment.

→ 8+ years as Technical Business Analyst / Scrum Master / Product Owner: stakeholder workshops, AS-IS mapping, documentation, turning findings into backlog items a team can actually execute.

Taking on AI Process Diagnostic engagements now. What's the one process quietly costing your team the most time? Let's find out if it's worth automating.

Python · LangGraph · Claude/Anthropic API · RAG · ChromaDB · sentence-transformers · SQL · Scikit-learn · TensorFlow/Keras · PyTorch · HuggingFace · Prompt Engineering · MLflow · Docker · GCP · Streamlit · ServiceNow

Langues

Anglais
Bilingue ou natif
Espagnol
Bilingue ou natif
Néerlandais
Capacité professionnelle limitée
Portugais
Capacité professionnelle complète

Préférences en matière de lieu de travail

Accepte de travailler sur site

Antwerp (jusqu’à 50 km)

Le Wagon
Data Science, Machine Learning and AI
janvier 2026 - mars 2026 (2 mois)
Tokyo, Japan
Bootcamp completed March 2026 — fully dedicated to my AI consultancy practice from April 2026. Data Science and AI Bootcamp — Science, Machine Learning & AI ▸ Python and its data libraries (Pandas, Numpy, Matplotlib) ▸ SQL ▸ Data ETL ▸ Data visualization ▸ Machine learning modelling (Scikit-learn) ▸ Deep learning modeling (Tensorflow-Keras) ▸ CI/CD/CT (Git, GitHub, MLflow) ▸ MLOps (Docker, Google Cloud Platform) ▸ Generative AI Application and Ethics
artificial intelligence Machine learning Data science
KRS Solutions
Founder & AI Business Consultant
juin 2023 - Aujourd'hui (3 ans et 1 mois)
Antwerp, Belgium
risk classification, stakeholder alignment.→ 8+ years as Technical Business Analyst / Scrum Master / Product Owner: stakeholder workshops, AS-IS mapping, documentation, turning findings into backlog items a team can actually execute.Building this through KRS Solutions, my consultancy based in Antwerp, serving Belgium & the EU. Taking on AI Process Diagnostic engagements now.What's the one process quietly costing your team the most time? Let's find out if it's worth automating.Stack: Python
• · LangGraph
◦ · Claude/Anthropic API
• · RAG
• · ChromaDB
• · sentence-transformers
◦ · SQL
◦ · Scikit-learn
◦ ·

I build the AI myself rather than just advising on it — solo-built an agentic RAG tool (LangGraph, Claude, ChromaDB) that runs full EU AI Act compliance audits end-to-end (full writeup in Projects below). When a diagnostic surfaces a regulated AI system, that same tool powers an optional compliance gap check inside the engagement.
Servicenow
Business Process Analyst - HAM (Consultant)
mai 2023 - octobre 2025 (2 ans et 5 mois)
Brussels, Belgium
Led two AI-driven digital transformation projects for the European Commission logistics department (ATOS/DIGIT). PROJECT 1 — Asset Tracking App (built from scratch) ▸ Stakeholder discovery across warehouse teams, IT infrastructure and suppliers ▸ Designed app mockups (Balsamiq), features, user roadmap and data architecture across multiple disconnected databases ▸ Automated asset credentialing via QR scanning app — credentials assigned automatically upon each device scan ▸ Validated REST API integrations using Swagger, JSON, Postman ▸ UX/UI prototyping: Balsamiq → Adobe XD → Figma PROJECT 2 — (EU scale) ▸ Mapped AS-IS workflows and redesigned TO-BE processes across 5 business cases and 11 sub-cases
▸ Defined roles, access rights and governance rules across a highly complex institution ▸ Designed RBAC including a management tool giving team leads real-time visibility of hardware/software with approval workflows ▸ Automated full hardware lifecycle: reception → storage → configuration → assignment → delivery ▸ Managed warehouse layout: pallet organisation, supplier reception and delivery sequencing ▸ Designed pre-delivery software installation workflow with automated expiry tracking ▸ Automated software license expiry alerts (SAM) — preventing service interruption for thousands of EC users ▸ Managed simultaneous onboarding for 400+ new users Result: two systems replacing manual processes — structured, automated and AI-ready for predictive analytics at EU scale.