I'm Niklas, an Associate ML Engineer. This is my personal repository containing some small side projects of mine. I'm passionate about building end-to-end AI applications.
My focus lies on Python, with some SQL mixed in. I have gathered some front-end experience in Python with Streamlit, and some back-end experience with FastAPI. Also, I am an avid advocate of using the markdown editor Obsidian.
I have studied a combination of Quantitative Finance & Computer Science at the Technical University of Munich (TUM). My master's thesis leveraged different Machine Learning models to forecast Corporate Earnings up to three years into the future, derive firms' implied cost of capital estimates from these forecasts, and finally formulate investment strategies that outperformed those based on financial analysts' earnings estimates.
I'm currently studying about MLOps / LLMOps and how to effectively utilize technologies such as Docker, CI/CD pipelines, Metaflow and MLflow to build fullstack AI applications.