Data & AI Systems Engineer
Saarland, Germany
I build data and AI systems with a focus on traceable data flows, clear access boundaries, and long-term evolution.
I build data and AI solutions at the intersection of architecture, operations, and governance. What matters to me is not only that a system works, but that its data flows, access model, and future evolution are understandable from the start.
I learn new technologies best by building with them, in my own projects and through self-hosting. That lets me evaluate not only what a tool promises, but also how it integrates and where its boundaries lie. Swimming and hiking provide the balance away from the screen.
These projects show how I implement enterprise-oriented data and AI architectures from integration through to operations. The focus is on clear data flows, controlled access, repeatable delivery, and hands-on exploration of new technologies.
Enterprise-oriented, self-hosted single-host lakehouse for retail data. It combines batch processing, open table formats, role-based data access, and analytics in a local environment.
Local agentic BI environment for controlled queries on modelled data. It shows how chat agents access modelled data through a semantic layer and which identity and authorization steps this requires.
Local MLOps workflow for a classification scenario. It combines training, tracking, batch inference, and monitoring to make model lifecycles tangible.
Reproducible Fabric analytics flow for financial transactions. It focuses on infrastructure as code, medallion processing, and the publication of analytics artefacts.
Designed an agentic AI solution to simulate and analyse sales data. Developed an end-to-end MLOps workflow in Azure Machine Learning for customer churn classification. Implemented a recommendation system for banking products using hybrid collaborative-filtering methods. Supported the migration from Azure Synapse Analytics to Microsoft Fabric.
Analysed and compared MLOps lifecycles across Azure, Google Cloud Platform, and AWS. Designed and implemented an on-premises lakehouse architecture. Evaluated and integrated Microsoft Fabric into the central technology portfolio.
Developed and evaluated linear models to estimate vehicle fuel consumption from simple sensor data. Then contributed to software development projects with a focus on front-end engineering.
Evaluated suitable NoSQL database solutions. Set up a Cassandra cluster across multiple Docker containers. Developed a data model for storing vehicle speed data. Integrated Cassandra with the existing backend architecture.
Current certifications complement my practical work.
If you would like to connect or simply get in touch, I would be glad to hear from you.