🧠 AI Engineer | LLM Systems · RAG Pipelines · Hybrid ML Architectures · NLP

I build AI systems that actually work in production — not just in notebooks.

With 8+ years of experience and an MSc in AI Engineering (Germany), I specialize in LLM pipelines, retrieval-augmented generation (RAG), and hybrid ML + LLM systems for complex real-world problems. I focus on challenges where standard models fail — noisy data, ambiguity, and edge-case-heavy environments.

My work goes beyond training models. I design systems that reason, self-correct, and remain reliable under real-world conditions, combining research-level depth with practical engineering to deliver scalable, production-ready AI solutions.

📊 Results That Speak
* ~89% accuracy on imbalanced real-world data using hybrid ML + LLM architecture
* Cohen’s Kappa = 0.811 (automated essay scoring, ensemble models)
* Macro-F1 = 0.81 on multilingual NLP tasks (SemEval 2025)
* Improved planning & resource allocation via time-series forecasting (Germany)

🚀 What I Can Build for You
1. LLM & GenAI Systems
* End-to-end RAG pipelines (retrieval, chunking, reranking)
* Document intelligence, QA systems, and AI assistants
* Prompt engineering + structured reasoning workflows
* AI agents and multi-step automation pipelines
* LLM evaluation: accuracy, robustness, and failure analysis

2. Machine Learning & NLP
* Classification, clustering, recommendation systems
* Multilingual NLP (BERT, RoBERTa, DeBERTa, Sentence Transformers)
* Sentiment analysis, topic modeling, entity extraction
* Feature engineering for noisy, imbalanced, low-resource data
* Ensemble and hybrid models for maximum performance

3. AI Engineering & Deployment
* End-to-end pipelines (raw data → production system)
* API deployment and interactive apps (Gradio)
* Hyperparameter tuning (Optuna)
* Scalable, reproducible ML systems

4. Data Science & Analytics
* EDA, statistical modeling, and predictive analytics
* Time-series forecasting and demand prediction
* Customer segmentation and behavior analysis
* A/B testing and experimental design

5. Explainable AI & Evaluation
* SHAP-based explainability and feature attribution
* Robustness testing, ambiguity handling, edge-case evaluation
* Model benchmarking and failure-case analysis

🔧 Tech Stack
1. Languages: Python · R · SQL

2. ML/NLP: HuggingFace · XGBoost · LightGBM · Scikit-learn · Optuna

3. Transformers: BERT · RoBERTa · DeBERTa · Sentence Transformers · MiniLM

4. LLMs: RAG pipelines · prompt engineering · evaluation frameworks

5. Explainability: SHAP

6. Data: Pandas · NumPy · feature engineering pipelines

7. Infra: API deployment · experiment tracking · reproducible pipelines

8. Tools: Jupyter · Git · Gradio · VS Code · Tableau

🎓 Education & Certifications
MSc Artificial Intelligence Engineering — Universität Passau

Associate Data Scientist in Python — DataCamp

Data Analyst in Python — DataCamp

Computational Statistics in R — Universität Passau

✅ Why Clients Choose Me
* Expertise in LLM + RAG systems (beyond basic ML)
* Strong experience with real-world noisy & imbalanced datasets
* Focus on robustness, evaluation, and failure handling
* Experience designing hybrid ML + LLM systems
* Ability to deliver production-ready AI solutions end-to-end
* Research-level depth with practical engineering mindset
* Clear, business-oriented communication
* Reliable, detail-oriented, deadline-driven

🤝 Let’s Work Together
If you need more than a model — if you need a reliable AI system that performs in production — I can help.
Let’s build AI solutions that deliver real, measurable impact.