24.02.2026 aktualisiert


100 % verfügbar
Data Scientist & Ökonom (PhD) | Machine Learning, Beratung & Kausalanalyse | Ex-Ministerium
Berlin, Deutschland
Weltweit
PhD in EconomicsÜber mich
Ich bin Data Scientist und promovierter Volkswirt mit über zehn Jahren Erfahrung in Datenanalyse, Ökonometrie, Machine Learning und strategischer Kommunikation. Meine Stärke liegt darin, komplexe Daten in klare, handlungsrelevante Erkenntnisse zu übersetzen.
Skills
A/B TestingAPIsKünstliche IntelligenzAmazon Web ServicesData AnalysisKünstliche Neurale NetzwerkeAutomatisierungBash ShellBusiness IntelligenceBig DataCloud ComputingContinuous IntegrationDatenvalidierungDatenbereinigungInformation Engineering
Programmierung & Datenverarbeitung
Python · R · SQL · VBA · NumPy · pandas · statsmodels · APIs · Flask · FastAPI · PySpark · MLflow · Docker · Linux · Bash · Data Cleaning · Data Validation · Reproducible Research
Machine Learning & KI
Machine Learning · scikit-learn · XGBoost · Random Forest · Gradient Boosting · Logistic Regression · SVM · Feature Engineering · Model Evaluation · Hyperparameter Tuning · Cross Validation · Model Deployment · Monitoring · Bias & Error Analysis
Deep Learning & KI-Systeme
Deep Learning · TensorFlow · Keras · PyTorch · Neural Networks · CNNs · Transfer Learning · Representation Learning · Model Explainability (SHAP, LIME) · Reliability & Robustness Testing
LLMs, NLP & RAG
Large Language Models (LLMs) · Prompt Engineering · Retrieval-Augmented Generation (RAG) · Hugging Face · Embeddings · Semantic Search · Reasoning Evaluation · Structured LLM Evaluation · Hallucination & Error Analysis · NLP Pipelines · NLTK · spaCy · Text Classification · Sentiment Analysis
Datenvisualisierung & Business Intelligence
Data Visualization · Matplotlib · Seaborn · Plotly · ggplot · Power BI · Tableau · Data Storytelling · Dashboard Design · KPI Definition · Stakeholder Reporting
Causal Analysis, Statistik & Ökonometrie
Statistical Modelling · Econometrics · Causal Analysis / Causal Inference · Difference-in-Differences (DiD) · Instrumental Variables (IV) · Propensity Score Matching · Regression Discontinuity · Panel Data Analysis · Time Series Analysis · Forecasting · Experimental Design · A/B Testing
Data Engineering, Big Data & Cloud
Data Pipelines · Data Warehousing · Distributed Computing · Spark · Databricks · Lakehouse Architectures · AWS (Cloud Practitioner zertifiziert) · Cloud Deployment · Streamlit · Scalable Analytics
MLOps & Workflow Automation
MLOps · MLflow · Docker · CI/CD · Workflow Orchestration · AWS Step Functions · Automated Model Evaluation · Experiment Tracking · Reproducible Pipelines
Web Scraping & Datenakquise
Web Scraping · Beautiful Soup · Data Extraction · API Integration · Structured & Unstructured Data Processing
Versionierung & Systeme
Git · GitHub · GitLab · CI/CD · Shell Scripting · Linux Environments
Business, Economics & Decision Support
Economic Modelling · Financial Modelling · Risk Analysis · Policy Evaluation · Data-Driven Decision Making · Strategic Communication · AI-supported Decision Systems
Erfahrung
Berufserfahrungmehr als 10 Jahre
Freelancer seit2024
Sprachen
DeutschMutterspracheEnglischverhandlungssicherFranzösischGrundkenntnisseSpanischgut
Projekthistorie
- Develop, evaluate and optimize advanced AI/LLM systems for economic decision-making, including benchmarking, reasoning assessment, reliability checks and error analysis.
- Improve model accuracy and domain-specific performance through prompt engineering, RAG/LLM optimization techniques and deep research rubrics.
- Build a predictive model identifying attrition risks among young customers and partner with the national banking association to design a data-driven strategy for long-term customer retention.
- Develop and evaluate data models (XGB) for automated customer targeting across 40 million accounts, including KPI design and impact measurement to optimize sales channels.
- Project in AI: Lung Disease Detection using X-Ray Images
- Develop deep learning and classic ML (SVM, XGB) models to detect lung diseases from 21,000 X-rays, achieving a 95% macro F1-score with the best TensorFlow transfer learning model.
- Deploy a Streamlit web app on AWS Cloud enabling real-time predictions and Grad-CAM visualizations from uploaded chest X-rays.
Zertifikate
Data Scientist
Université Paris I Panthéon-Sorbonne2024