I am a Data Scientist specializing in building and fine-tuning Large Language Models (LLMs). My focus is on using advanced AI to solve complex and meaningful problems.
With a background combining quantitative analysis from engineering and economics, I bring a unique and rigorous approach to my work. I am fluent in English and German and am excited to connect with teams building impactful technology in Europe.
- Description: An end-to-end Information Extraction system using models like GPT-4 and LLaMA, achieving 80% accuracy.
- ➡️ View Repository
- Description: A research project where I boosted the LLaMA-3.1 model's performance by 15% through hands-on LoRA fine-tuning.
- ➡️ View Repository
- Description: Retrieval-augmented PDF Q&A API using FastAPI, Gemini 1.5 Flash, and Embedding-001 (768-d) with PostgreSQL + pgvector (IVFFLAT) for cosine similarity, Redis caching, LangGraph workflows, and Docker Compose for local dev.
- ➡️ View Repository
- Description: Power BI sales dashboard on Azure SQL Database (provisioned via Terraform IaC) using a Kaggle dataset—Power Query for ETL, DAX for measures, secure firewall allowlisting, and direct SQL connections for scheduled/near-real-time refresh.
- ➡️ View Repository