At The HEINEKEN Company, I lead AI and machine learning initiatives across 30+ Operating Companies and manage a team of five data scientists and analysts. I set technical direction across agentic AI, predictive modelling, generative AI and analytics.
I architected and productionized BORIS, an Azure OpenAI agent for security operations serving 20,000 users. Its LangGraph and RAG architecture automates alert triage, incident summaries and threat enrichment, reducing analyst workload by 40% and enabling three times the incident throughput.
I built BREWCAST AI, a Databricks-based commercial and supply chain intelligence platform combining forecasting, sales, inventory and promotion analytics with RAG. It reduced commercial analysis time by 60% and improved inventory planning efficiency by 20%.
Earlier, at Viewture I built YouTube view forecasting models and analyzed thumbnail experiments; at Ejam I developed campaign models, advertising tests and a recommender. In my SheetPilot AI project, I built a conversational Excel agent with LangGraph, Gemini, FastAPI and Streamlit.

