Machine Learning · Deep Learning · AI Engineering
I enjoy developing reliable AI systems that transform complex data into practical, well-evaluated solutions.
My background combines applied mathematics, computer science and practical experience with machine learning applications in the energy and meteorological domains. I am particularly interested in reliable AI systems that combine machine learning with practical applications and systematic, measurable evaluation.
For me, good AI engineering is about understanding why a system works—not just that it works. The real challenge is not building a model that performs well in theory, but transforming it into a practical AI system through reliable software design, systematic evaluation and reproducibility.
An unnecessarily thorough investigation into whether the card game Pazaak cheats.
The project combines computer vision, automated gameplay extraction, Monte Carlo simulation, dynamic programming and statistical hypothesis testing. It includes a complete scientific investigation report covering the methodology, experiments and conclusions.
Focus: Computer Vision · Statistical Analysis · Scientific Documentation
Technologies: Python · OpenCV · NumPy · SciPy
📄 Read the full investigation report
A Retrieval-Augmented Generation system for context-aware customer-support applications.
The project includes document preprocessing, embeddings, FAISS-based retrieval, contextual reranking, safe response handling, a FastAPI interface, Docker support and automated tests. The current focus is the systematic evaluation and continued improvement of retrieval quality.
Focus: Retrieval-Augmented Generation · AI Engineering · Evaluation
Technologies: Python · FAISS · Sentence Transformers · FastAPI · Docker · pytest
My experience spans the complete AI development lifecycle—from data acquisition and preprocessing to model development, evaluation and deployment.
Areas I particularly enjoy working in include:
- Machine-learning forecasting for energy-market applications
- Multimodal deep learning for weather forecasting
- Retrieval-Augmented Generation systems
- Computer Vision pipelines for automated data extraction
- Scientific evaluation and technical documentation
Languages Python · Java · C · C#
Machine Learning PyTorch · TensorFlow · scikit-learn · AutoGluon
Computer Vision OpenCV
Data Pandas · NumPy · Xarray · Zarr · Dask
AI Engineering RAG · FAISS · Sentence Transformers · FastAPI
Tools Git · Docker · Linux · pytest · Jupyter
Cloud AWS S3