Featured
Something cool to show off
Chatbot integration coming soon
What I'm Up To Now
Transforming business processes through AI Engineering
Data Scientist /
AI Engineer Current
Jan 2024 – Present
Data Science Lab, Amsterdam
Consultant at a prominent AI consultancy firm in the capital of the Netherlands. Helped lead the NLP module — a self-initiated team of five consultants specializing in advanced AI techniques for text processing. The module was promoted to an official chapter, becoming one of the four pillars of the organization alongside data engineering, data science, and strategy.
Built an elaborate technical foundation for working with LLMs, creating a modular repository covering techniques from query transformation to automatic advanced chunking for chatbot integration. Organized and led a comprehensive NLP training day for the entire technical team.
Responsible for six complete proposal processes — speaking with prospects, determining technical scope, drafting agile proposals, and presenting. Converted 6 out of 7 prospects into paying clients (€25K–€103K per project).
The NLP module was promoted in 2025 to an official chapter, becoming one of the four pillars of the organization alongside data engineering, data science, and strategy.
Projects at Data Science Lab
Customer Service Analysis
AI Engineer Current
Jan 2026 – PresentIndepender, Hilversum
Building and maintaining scalable AI pipelines to classify and analyze tens of thousands of customer care conversations using speech-to-text systems and large language models.
Independer is a leading Dutch online comparison platform that helps consumers make informed decisions on financial products such as insurance, mortgages, and energy contracts.
These pipelines operate on both large-scale historical datasets and daily streaming data, enabling continuous monitoring, insight extraction, and quality analysis.
In parallel, developing automation solutions for other business workflows, including contract document extraction and structured data retrieval, leveraging modern NLP and information extraction techniques to reduce manual processing and improve operational efficiency.
HR Chatbot
AI Engineer
2025Fokker Services Group, Amsterdam
Developed a chatbot for Fokker's HR department to handle incoming questions, integrating documents with tables, infographics, and images using OCR & advanced chunking for RAG applications.
- Preprocessing script to extract tables, table images, and infographics from documents
- RAG model combining state-of-the-art retrieval techniques
- Evaluation dataset with multiple indicators (text, tables, infographics, conversation history, multi-hop questions, guardrails)
- Structured evaluation with MLFlow for retriever and generator
- Guardrails against off-topic answers, hallucinations, prompt injection
- Memory management, logging, monitoring, latency optimization
- Multilingual effectiveness & front-end development
Git, Azure DevOps (Scrum, CI/CD), RAG, MLFlow, Azure Cloud, LLMs, Pydantic, MCP, OCR
News Summarizer
AI Engineer
2025Fokker Services Group, Amsterdam
Built a news summarizer that automatically scrapes and summarizes articles from predefined websites, with dynamic parameters for company name, websites, time interval, and language.
- Dynamic web page search via headless browser based on 3 parameters
- Web scraping with BeautifulSoup
- Post-processing with keyword filtering & LLM filtering
- Multi-language report generation
- Logging and monitoring traces for optimization
- Front-end development
Git, BeautifulSoup, headless browser, MLflow, Azure Cloud, LLMs, Pydantic, MCP
Internal Chatbot
AI Engineer
2025Data Science Lab, Amsterdam
Developed an internal chatbot connected to SharePoint via MCP, using RAG for document navigation and Q&A on HR documents, project documentation, and onboarding support.
- Hybrid search (vector search + BM25), reranking, reciprocal rank fusion, metadata filtering
- Preprocessing script for documents with ambiguous titles
- Evaluation dataset & structured evaluation with MLFlow
- Guardrails, memory management
- CI/CD pipeline in Azure, deployment via Azure & Docker
- Logging and monitoring conversation traces in production
- Secure front-end development
Git, Azure DevOps (Scrum, CI/CD), RAG, MLFlow, Azure Cloud, Docker, LLMs, Pydantic, MCP
AI Interviewer
AI Engineer
2025WUA, Amsterdam
Created an AI interviewer replacing static questionnaires with dynamic LLM-driven interviews for clients like ING, Rabobank, and Vodafone.
- Preprocessing script & API development with FastAPI
- OTAP segmentation & CI/CD pipeline in Azure
- API testing & scaling concurrent request handling
- Confusion matrix optimization for result analysis
- 'Question answer' LLM model based on industry techniques
- Memory management & guardrails implementation
Git, Azure DevOps (Scrum, CI/CD), FastAPI, Postman, Azure Cloud, Docker, LLMs
Matching Tool
AI Engineer
2025Zorginstituut Nederland, Diemen
Created an LLM-based matching tool to automate data verification for the KIK-V program, replacing manual domain expert review.
- Preprocessing script for ontologies
- Workflow to match external data to correct columns via metadata filtering
- Matching LLM model based on industry techniques
- Confusion matrix optimization for result analysis
- Front-end for end-users managing the data station
Git, Azure DevOps (Scrum, CI/CD), Azure Cloud, Docker, LLMs, Pydantic
Semantic Web Engineering
Data Engineer
2025Zorginstituut Nederland, Diemen
Semantic web data engineer on the KIK-V program, developing SPARQL queries and creating training programs for RDF and ontology tooling.
- Translating provider data into RDF format as triples per the ontology
- Python tests for SPARQL queries to compare outcomes
- SPARQL queries from functional analyses into technically correct queries
- Ontology visualization with Protégé
- Graph databases (GraphDB, Fuseki) for data loading & querying
- Training program with extensive documentation for new employees
Git, Gitlab (Scrum, CI/CD), SPARQL, GraphDB, Fuseki, Azure Cloud, Protégé
AI Engineer
AI Engineer
2024Athlon, Almere
Developed NLP and LLM-based text analysis for customer reviews, including topic modeling and a production API. Created a company-wide dashboard in Power BI.
- Preprocessing script & dataset analysis for correlations
- "Topic Modeling" model based on machine learning (BERT)
- Training ML model to estimate topic and subtopics of reviews
- API development for production classification
- Dashboard for relevant review statistics
Git, Azure DevOps (Scrum, CI/CD), Snowflake, Python, REST APIs, Azure Cloud, Power BI, Qlik Sense, LLMs, Pydantic
Education
Tilburg University
2021 – 2023Pre-Master & Master in Data Science & Society
Multidisciplinary program focused on applying mathematics, computer science, and statistics to improve business processes.
Master's Research Paper, TiU International Office: "Classifying Earthquake Damage in Nepal: A Comparative Study of Tree-Based Algorithms and Deep Learning."
Python, R, LaTeX, Power BI, SQL
Amsterdam University of Applied Sciences
2016 – 2021Bachelor in International Business & Languages
Academic and practice-oriented program at the intersection of economics, management, and cultural relations.
Graduation Project for Wallbox: A qualitative study on which market entry strategy Wallbox should implement to successfully enter the Benelux market.