Working Student at ASML | M.Sc. AI at BTU Cottbus-Senftenberg
I specialize in industrializing AI systems—bridging the gap between research models and production tools. My work focuses on building scalable Computer Vision pipelines, Agentic AI workflows, and interactive Data Engineering platforms that process engineers actually use.
- 🤖 Agentic AI Frameworks: Architecting modular multi-agent systems using LangChain/Langflow and RAG for complex reasoning tasks.
- 🏭 Industrial ML: Developing Deviation Predictors for semiconductor manufacturing using SVD completion and uncertainty quantification.
- 📊 High-Res Analytics: Building interactive Dash/Plotly platforms to visualize 1024x1024 interferometry data, reducing analysis time by 40%.
- ⚙️ Data Engineering: Designing end-to-end ETL pipelines and automated quality checks on Azure.
Agentic AI & GenAI
- Orchestration: LangChain, Langflow, Multi-Agent Systems
- RAG: Vector Databases (Qdrant, ChromaDB), Context-aware retrieval
- APIs: OpenAI, Anthropic, Hugging Face
Machine Learning & Vision
- Core: PyTorch, Transformers, scikit-learn
- Vision: OpenCV, scikit-image, Surface Reconstruction
- Methods: Structural Pruning (ViTs), SVD, Anomaly Detection
Data & Engineering
- Stack: Python (FastAPI, Asyncio), Pandas, Parquet
- Cloud/Ops: Docker, Git, Azure Data Factory, CI/CD
- Visualization: Plotly, Dash
- Scaling Agentic Workflows for reliable enterprise applications.
- Structural Pruning of Vision Transformers (Research Paper under review).
- Industrial MLOps: Making ML robust enough for semiconductor production lines.
