I'm a Data Scientist Consultant at Orange Business, where I build production-grade AI systems. I recently shipped a RAG pipeline and I'm currently deep into agentic architectures: autonomous systems that reason, plan, and act.
At Orange Business, I work on applied AI projects at enterprise scale, from designing retrieval-augmented generation pipelines to building multi-step agent workflows. I care about turning research-grade AI into robust, deployable systems.
My background includes:
- 🤖 Agentic systems - LLM-driven workflows that reason and act autonomously
- 🔎 RAG pipelines - semantic retrieval, vector stores, document processing
- 🔤 NLP at scale - keyword generation, semantic search engines, dense vector representations
- ☁️ Cloud-native deployment - Python apps, APIs, CI/CD on AWS and Azure
Before Orange Business, I spent 4 years as an NLP engineer at Verbolia, a Belgian AI-SEO startup that auto-generates optimized landing pages for major e-commerce brands. That's where I built my foundation in semantic search, keyword systems, and production NLP pipelines.
In between, I took a year off in New Zealand on a Working Holiday Visa, one of the better decisions I've made.
I then joined BeCode as a Data Science & AI coach, guiding career changers through machine learning, cloud, and MLOps.
I believe learning by doing is the most effective path. My goal as a coach was to bridge the gap between theory and practice, encouraging students to build real-world solutions, write clean code, and understand the why behind the tools. That mindset still shapes how I work today.
Currently exploring:
- Agentic frameworks - multi-agent orchestration, tool use, memory systems
- LLM evaluation - measuring and improving agent reliability
- MLOps - making AI systems production-ready at enterprise scale