📍 Based in Bonn, Germany (Open to Relocation / Hybrid)
💼 Seeking Junior/Entry Positions in Data Analytics, Data Science & Corporate Strategy

Data-driven professional bridging the gap between rigorous econometric modeling and production-ready data pipelines. Experienced in containerizing Python tracking workflows, structuring high-performance relational architectures (O(log n) efficiency), and deploying corporate BI and antitrust simulation layers.

⚡ Languages & Core Analytics:

⚙️ Data Infrastructure & Tooling:

• Repository: merger-simulation-hhi-analyzer
• Tech Stack: Python, Google BigQuery, Streamlit, Plotly, Pydantic, GitHub Actions
• The Architecture: An institutional-grade Python toolkit and web dashboard automating the U.S. DOJ Horizontal Merger Guidelines. Engineered strict Pydantic data models, dynamic Plotly visual threshold gauges, and a native SQL pipeline to execute concentration simulations (HHI) directly against Google Cloud BigQuery.

• Repository: retail-media-clv-optimizer
• Tech Stack: Python, Google Cloud BigQuery, Lifetimes, PyTest, CI/CD, Power BI
• The Architecture: Built an end-to-end data lake ingestion track streaming customer matrices into BigQuery. Implemented parallel mathematical tracking loops utilizing probabilistic BG/NBD and Gamma-Gamma models to project 12-month customer horizons with an automated 98.67% testing gate.

• Repository: equity-impact-predictor
• Tech Stack: Python, LightGBM, Streamlit, Plotly, HuggingFace (Zero-Shot NLP), SHAP
• The Architecture: An institutional-grade quantitative monitor forecasting short-term Cumulative Abnormal Returns (CAR) from market news. Utilizes BART-Large-MNLI for sentiment classification and LightGBM for predictive modeling. Features a full decision-terminal UI with Explainable AI (SHAP) breakdowns and real-time market regime/drift tracking.

• Repository: invoice-llm-pipeline
• Tech Stack: Python, GenAI/NLP Frameworks, Structured JSON parsing, JSONL Curation
• The Architecture: Engineered a token-aware context window wrapper using sliding character segmentations to handle complex, unstructured billing texts. Enforced strict schema validation constraints, structuring outputs into formatted .jsonl tracks ready for downstream SFT fine-tuning loops.

• Repository: pricing-ab-simulator
• Tech Stack: Python, SciPy, NumPy, Matplotlib, Data Architecture
• The Architecture: Designed an automated pricing experiment pipeline running continuous hypothesis testing routines to evaluate localized price elasticity. Integrated programmatic power analysis checks to calculate optimal sample boundaries, protecting models from Type-I/II execution errors.