Full-stack Developer in Kiel, Schleswig-Holstein, Germany

I developed a smart home system from the ground up.

For system administration, I built a React SPA with a PHP back end, providing a REST API. This API was then used by hardware components like lamps, switches, and laser barriers that I built based on the ESP8266 and programmed with Arduino. I soldered them together from PCBs, designed in Eagle, and put them into 3D printed cases designed in OpenCAD.

In addition, the system used Raspberry Pis to monitor rooms and detect people, which triggered an event that was used to automatically keep the lights (Philips Hue) on only in rooms with people.

For my Bachelor thesis project, I compared machine learning algorithms in the context of the game EvoMan.

This Python project used four machine learning algorithms, NeuroEvolution of Augmenting Topologies, Deep Q-Network, Advantage Actor-Critic, and REINFORCE, to compare their performance in the context of the open-source game EvoMan. First, the agents were trained against a pre-programmed AI, and then their performance was measured against that AI and against each other.

Developed and deployed an NFT smart contract on Goerli using zero-knowledge proofs to validate the NTF's minting. This project was a proof of concept in the context of a Udacity Nanodegree program in Blockchain development and could be further expanded into a complete Web3 marketplace (e.g., real estate).