Stefan Schneider

Large railway project to apply future technologies to railway operation, from research to implementation. My team focuses on improving and dynamically adjusting railway schedules with deep reinforcement learning, leading to higher capacity, less delays, and greener transportation.

Research project on machine learning for autonomous coordination in 5G and beyond.

• Developed distributed and robust machine learning and deep reinforcement learning approaches for 5G networks and beyond
• Collaborated with researchers from Huawei Munich

My own research project on self-learning network management through machine learning.

• Planned, applied for, and received 100K grant to launch my own IT research project
• Recruited and leading 4 student assistants & collaborating with researchers from Huawei
• Combining networking and machine learning to tackle real-world problems

Innovation action with 18 European partners from industry and academia.

• Designed and developed cloud-native smart manufacturing scenario and deployed it on Kubernetes
• Implemented various tools to simplify and accelerate 5G service development

Design and development of a prototyping platform congestion control and messaging of an ETSI ITS vehicular networking stack. Publication and presentation of results and live demo at the 2016 IEEE Vehicular Networking Conference.

Competing in the ongoing LLM-agent challenge on real-world travel planning at IJCAI 2026, building an agentic pipeline around an open LLM (Qwen). Currently ranked 4th on the leaderboard.

Hands-on exploration of the LLM and agent stack: fine-tuned an encoder-decoder model (Longformer) for long-form question answering, built a Q&A demo app, and wrote a custom MCP server. Write-ups on my blog.

A simple Django web app using a pretrained PyTorch DenseNet model for classifying uploaded images. Deployed on Heroku.

Personal blog for documenting things that I learned or want to learn. Mostly related to machine learning, networking, or technology in general.

Progressive Web App (PWA) to securely and easily track food, activities, and personal well being. Visualize tracked entries to understand what has a positive impact and what doesn’t. Implemented with Django and Bootstrap. Deployed on Heroku.

Progressive Web App (PWA) to keep track of new ideas, organize them with tags, and add notes. Implemented with Django and Bootstrap. Tested and deployed automatically on Heroku.

Website with up-to-date information and data on the COVID-19 spread. Visualization and analysis of trends worldwide and, in particular, in Germany.

Contributor to Ray RLlib, a widely used open-source library for scalable deep reinforcement learning.

Open, Gymnasium-based platform for reinforcement learning research in wireless mobile networks. Published at IEEE/IFIP NOMS and used by researchers and students.