Lead Machine Learning Engineer
I build data science systems that make it into production and stay useful there.
Based in Switzerland
MSc Engineering, Data Science
ti&m AG
Building and running machine learning projects, owning their architecture and MLOps. Responsible for technical recruiting.
Universität Luzern
Teaching two courses in the Master in Economics and Management: Data Science Toolkits and Architectures, and Unsupervised Machine Learning.
ti&m AG
Ran data science projects from the first analysis through to running them in production. Designed and tuned retrieval-augmented generation architectures, in the cloud and on-premises.
ti&m AG
Built and maintained software with machine learning in it, mostly for banking and finance clients.
Schindler Group
Built a web application used across the group to automate testing of elevator control software. Front end, back end, and the test suite around it.
Hochschule Luzern
Native iOS and Android work on the university mobile apps, including maintenance and App Store releases.
SUISA · · Senior ML Engineer
Transcribes incoming call recordings and hands them to customer service as structured meeting notes. Speakers are separated and labeled, calls are summarized, and staff can see the status of each recording.
Innosuisse · · Senior ML Engineer
An internal GPT-based chat assistant built inside the client Azure environment and fitted to their existing processes and IT. Designed, built, and deployed under strict data-protection requirements.
Raiffeisen Schweiz · · Solution Architect
The internal AI platform at Raiffeisen, now carrying more than six use cases and open to the whole bank. Built on a RAG foundation with an evaluation pipeline, answers that cite their sources, and automatic hyperparameter tuning.
Government Agency · · Senior ML Engineer
An on-premise RAG chatbot for sensitive government data. Fine-tuned open-source LLMs with hybrid retrieval on Elasticsearch, running entirely inside an air-gapped network.
Hochschule Luzern · · Data Scientist
Next-purchase prediction for over 50'000 customers, learned from more than 13 million transactions. Transformer-based sequential models, with contrastive learning to stop the representations collapsing.
Raiffeisen Schweiz · · Senior ML Engineer
Compiles financial reports with LLMs, drafting and then refining them. Wired together over REST across Spring Boot and Python services.
Wrote the tender documents for the public procurement of a secure AI platform for Swiss municipalities and cities. Defined the architecture and operating requirements for a multi-tenant private-cloud or on-premise system handling highly sensitive data, serving LLM, translation, transcription, and search under Swiss compliance rules.
A multilingual RAG assistant on the public SUISA site. Built the chatbot front end and owned its integration and delivery, down to the streaming API behind it.
A customer platform for a consumer-credit bank, built as microservices. React and Next.js on the front, Spring Boot services behind it, Kafka for events and Camunda for workflows.
Detects food trends in social media posts using named entity recognition, topic detection, and time series forecasting.
Reads fencing technique from video using 2D pose estimation. Published as an IEEE conference paper.
Web application for automating tests of elevator control software, with JWT authentication against Active Directory.
Predicts temperature-related elevator door failures by mining log messages, using Apache Spark and deep learning.
Native Android and iOS apps for university operations and student services.
Hochschule Luzern · —
Hochschule Luzern · —