25.06.2026 aktualisiert
100 % verfügbar
Senior Data Engineer / AI Specialist
Murnau, Deutschland
Deutschland
Computer Science (Informatik) DiplomaSkills
Künstliche IntelligenzBig DataData ArchitectureMachine LearningData ScienceGenerative AI
With a focus on AI, data science and data engineering, I collected over 17 years' professional experience. I have developed AI and machine learning systems and was lead architect for companies in the energy, telecommunications and automotive sectors, as well as for small start-ups in fields such as medical technology. I also worked in research and gave talks on AI and statistics (see the end of this CV).
Aside from AI, my greatest strengths lie in data and software engineering & architecture. I love working on more than just models and have a strong focus to bring a project into production and gather feedback from stakeholders.
I really enjoy working with younger colleagues, both to pass on my knowledge and to learn from them.
Languages: German (mother tongue), English (professional)
Erfahrung
Berufserfahrungmehr als 10 Jahre
Freelancer seit2023
Sprachen
DeutschMutterspracheEnglischverhandlungssicher
Projekthistorie
A) Internal AI chatbot for a major telecommunications company in Munich: I designed and developed an AI chatbot, including a pre-processing pipeline for internal knowledge of a RAG system. Crafting prompts for all layers and features of the architecture. Improving the results of semantic search via hypothetical document embeddings. Challenge: Due to data privacy restrictions, the chatbot is running On-Premise with Open Source language models (now: Gemma4). I evaluated different language and embedding models and made it operable via Ollama.
B) ML Ops: I set up a MLFlow Server and migrated existing analytics / ML models to run on the new infrastructure, including migrating Jupyter notebooks to plain Python code and configuration.
Skills: Open Source Language Models, RAG, Ollama, OpenWebUI, Python, MLFlow, Podman
Right now I'm working on my own project developing AI agents controlling the browser resp websites using Stagehand (NodeJS / Typscript), Django and running on Google GCP via Terraform.
For different Web analytics use cases and sources I implemented a layered data architecture (Bronze, Silver, Gold “medaillon” architecture) which incrementally and progressively improves the structure and quality of data as it flows through each layer.
The final pipelines are deployed on Databricks (GCP), but the Python code is implemented and structured so it can run and be tested completely w/o Databricks e.g. with Github actions.
Delta tables are used so data versions can be rolled back. Updates based on data partitions are implemented, so data can be easily updated.
In another project I was working on Microservices inside a data streaming pipeline with Kafka and Python.
The final pipelines are deployed on Databricks (GCP), but the Python code is implemented and structured so it can run and be tested completely w/o Databricks e.g. with Github actions.
Delta tables are used so data versions can be rolled back. Updates based on data partitions are implemented, so data can be easily updated.
In another project I was working on Microservices inside a data streaming pipeline with Kafka and Python.