10.11.2026 aktualisiert


60 % teilweise verfügbar
Senior Data Scientist / ML Engineer
München, Deutschland
Deutschland
PhD Statistik und Machine LearningÜber mich
Senior Data Scientist with a strong academic foundation and years of industry experience — a combination that helps me ask the right questions, design solutions that truly fit the problem, and implement them in end-to-end systems.
Skills
AI AgentsBedrockAzure DatabricksKünstliche IntelligenzDatenvalidierungETLData IntelligenceMachine LearningStatistische AuswertungenAzure Machine LearningTransformerSimulationenComputational StatisticsZeitreihenanalyseBayes
Methoden: Machine Learning, Data Science, Statistik, Time Series Forecasting, Demand Forecasting, Bayesian Statistics, Uncertainty Quantification, NLP, Deep Learning, Clustering, Constraint Programming, Simulation
AI: Generative AI, LLM, RAG, AI Agents, Multi-Agent-Systeme, Prompt Engineering, LLM Fine-Tuning, LLM-Evaluation
Programmierung: Python, SQL, R, Julia
Libraries: pandas, scikit-learn, PyTorch, Hugging Face, LightGBM, PyMC, PySpark, LangGraph, OR-Tools, SHAP
Cloud & MLOps: Azure, Databricks, GCP, Palantir Foundry, Airflow, MLflow, Docker, CI/CD, Langfuse, LangSmith, ETL, Dashboards
Arbeitsweise: Scrum, Jira, Teamleitung
Erfahrung
Berufserfahrung5 - 10 Jahre
Freelancer seit2017
Sprachen
DeutschMutterspracheEnglischverhandlungssicherSchwedischGrundkenntnisse
Projekthistorie
Freelance AI Engineer for a large Consultancy Company.
Projects include:
* RAG solution to ask questions about a large corpus of contract documents in the automotive sector, using agentic frameworks, structured extraction of information, OpenAI stack and state-of-the-art RAG frameworks. Frontend deployed in databricks.
* Building up a data platform for a large german telecommunication company in Palantir Foundry, including data ontologies, data ingestion pipelines, visualisation dashboards and data entry forms.
* Using LLM's to match messy data in structured databases and clean messy datasets.
Coding in Python and Typescripty, using best coding practices
Projects include:
* RAG solution to ask questions about a large corpus of contract documents in the automotive sector, using agentic frameworks, structured extraction of information, OpenAI stack and state-of-the-art RAG frameworks. Frontend deployed in databricks.
* Building up a data platform for a large german telecommunication company in Palantir Foundry, including data ontologies, data ingestion pipelines, visualisation dashboards and data entry forms.
* Using LLM's to match messy data in structured databases and clean messy datasets.
Coding in Python and Typescripty, using best coding practices
AI Engineer and Data Scientist for the Scale-Up Company ViaEurope.
Several Projects that are currently in production:
* AI powered web-crawler to collect critical logistics data from a myriad of airline websites using LLMs and agentic frameworks (Langchain, Langsmith).
* Automated Extraction of structured data from PDFs, emails and Scanned Documents using OCR and LLM's.
* Machine Learning forecasts of Parcel arrival times based on very heterogenous data (flight data, parcel data, data from web scraping).
* Planning of Trucks and Workforce based on Parcel forecasts using Bayesian Modeling on the overall volume.
* Software Engineering Tasks, such as Building API endpoints, setting up Airflow, deploying solutions with docker and docker compose.
Several Projects that are currently in production:
* AI powered web-crawler to collect critical logistics data from a myriad of airline websites using LLMs and agentic frameworks (Langchain, Langsmith).
* Automated Extraction of structured data from PDFs, emails and Scanned Documents using OCR and LLM's.
* Machine Learning forecasts of Parcel arrival times based on very heterogenous data (flight data, parcel data, data from web scraping).
* Planning of Trucks and Workforce based on Parcel forecasts using Bayesian Modeling on the overall volume.
* Software Engineering Tasks, such as Building API endpoints, setting up Airflow, deploying solutions with docker and docker compose.
Several Coding Projects for a swedish insurance consultancy, mostly in R and Python.
Includes:
* Optimizing Code using Parallelization
* Conversion of Legacy Code into Python
* Development of Statistical Metrics in the Insurance Space to messaure Bonds performance.
Includes:
* Optimizing Code using Parallelization
* Conversion of Legacy Code into Python
* Development of Statistical Metrics in the Insurance Space to messaure Bonds performance.