Felix Schelling

AI Engineer and Software Developer with 5+ years of experience specializing in reinforcement learning applications for industrial optimization problems. Proven track record of developing innovative solutions for complex engineering challenges. Combining technical expertise with entrepreneurial experience to bridge the gap between AI research and practical industrial applications.

Professional Experience

2024–Present | Research Assistant at TH Rosenheim

Designed and implemented an open source reinforcement learning framework for solving job shop scheduling problems
Trained state-of-the-art models in a distributed training environment
Integrated trained models into a real-world factory lab environment

Academic Research Reinforcement Learning Dash SQL Optuna Sphinx Pytest

2023–Present | Co-Founder & CTO at Advanced Additive

Prototyped a reinforcement learning-based algorithm for pathfinding under domain-specific constraints for 3D printing slicers
Designed a heuristic algorithm to optimize the density of 3D printed parts utilizing a variable extrusion width
Deployed a scalable containerized web app with user management as an alpha release for customers
Managed diverse technical collaborators, including freelancers and students, for the startup

Project Management Reinforcement Learning Numeric Optimization Unsupervised Learning SQL FastAPI Docker Redis Celery Dask CI/CD

2021–2023 | Career Break

Completed a two-year bike journey from Argentina to Alaska

2019–2021 | Software Engineer at Krones GmbH

Designed and implemented a reinforcement learning model for solving a decision problem for a delta robot in a palletizing process
Integrated the model into a web app
Validated and documented the solution

Software Design Reinforcement Learning Python PyTorch gRPC Scrum Git

Education

2020 | Bachelor's Degree in Mechanical Engineering at TH Rosenheim

Completed thesis on automated parameter generation for a delta robot using a reinforcement learning model at Krones — Grade 1.1
Developed machine learning models for time series prediction of household oven usage in a voluntary project at BSH (Bosch/Siemens) — Grade 1.0