Beschreibung

Most logistics problems look like chaos from the outside. I see structure.

I am a self-taught Operations Research scientist specialising in vehicle routing, supply chain optimisation, and real-time decision systems. My work sits at the intersection of advanced mathematics and practical engineering. I build models that do not just look good on paper but actually run in production.

Recently, I developed a Deep Reinforcement Learning agent for real-time VRP that improved delivery efficiency by 15%. Not by tweaking a spreadsheet, but by rethinking how routing decisions get made dynamically, at scale.

What makes me different: I am not a black-box consultant. I explain every model, every trade-off, and every assumption in plain language so you understand what you are buying and why it works. If you are a logistics or supply chain company sitting on an inefficiency you cannot quite name, that is exactly the kind of problem I am built for.

Tools: Python, Gurobi, Scikit-learn, SQL, Excel. Timeline: fast. Communication: clear.

Branchenexpertise

Sprachen

Englisch
Muttersprachlich oder zweisprachig
Deutsch
Konversationssicher

Arbeitsortpräferenzen

Nur remote

Führt Projekte hauptsächlich remote aus

Deutsche Post DHL Group
Operations Research Scientist
November 2025 - April 2026 (5 Monate)
Berlin, Germany
• • Built and scaled predictive maintenance models using Gradient Boosted Trees (XGBoost/LightGBM) to forecast vehicle downtime, saving €2M+ in annual maintenance overhead.
• • Developed hybrid ML-Optimization frameworks by integrating predictive demand models with combinatorial solvers, reducing forecasting error-driven costs by 12%.
• • Engineered Deep Reinforcement Learning (DRL) agents to solve dynamic vehicle routing problems, improving real-time adaptability to traffic fluctuations and increasing on-time delivery by 15%.
• • Leveraged unsupervised learning and clustering to optimize warehouse zone-skipping strategies, resulting in a 20% reduction in manual handling touches.
• • Key Tools: Python, SQL, PyTorch, TensorFlow, Statsmodels, Gurobi, Git, Cloud Platforms, PuLP, GeoPandas, Shapely
Afrikiko GmbH
Data Scientist
März 2022 - November 2025 (3 Jahre und 8 Monate)
Berlin, Germany
• • Designed and deployed machine learning models for classification, clustering, and recommendation systems using Python and ML frameworks, improving predictive accuracy by 18%.
• • Built scalable ETL pipelines and automated workflows that supported reliable data ingestion and real-time insights, contributing to robust model deployment.
• • Conducted time series modeling and statistical analysis to forecast demand and identify optimization opportunities, aligning with production pipeline improvements.
• • Led feature engineering and model optimization for production use on cloud platforms (AWS/GCP), ensuring scalable machine learning implementations.
• • Collaborated with cross-functional teams to translate customer requirements into technical solutions, reducing operational costs and enhancing overall model performance.
• • Key Tools: SQL, BigQuery, Python, Power BI, HuggingFace, Cloud Platforms
Electric Mobility Concepts GmbH
Machine Learning Engineer
September 2020 - Dezember 2021 (1 Jahr und 3 Monate)
Berlin, Germany
• • Developed and validated supervised and unsupervised ML models using Python and frameworks like TensorFlow and Scikit-learn to optimize fleet usage and customer demand prediction.
• • Supported MLOps workflows including model monitoring, version control with Git, and reproducible experiments to ensure a smooth ML pipeline.
• • Key Tools: Python, Scikit-learn, SQL, APIs, ML Pipelines, XGBoost, PyTorch