Constantin Dißelkamp

Founder, machine learning engineer, and physicist in Bonn, Germany.

• X: consdi
• GitHub: disselkamp
• LinkedIn: disselkamp-constantin
• Email: constantin@disselkamp.co
• GPG fingerprint: 98EA 1A6A 856B BB47 5587 998A 4B92 18E1 77FF 119D

Experience

• 2023–present · Founder and CEO, pagent, Bonn. AI-powered website experimentation and personalization; backed by High-Tech Gründerfonds.
• 2020–2023 · Machine Learning Engineer and Team Lead, Trality, Vienna. Machine-learning systems for algorithmic trading.
• 2017 · Guest scientist, Forschungszentrum Jülich (PGI-1). Contributed spin-transfer-torque methods to Spirit on GitHub.

Selected work

• 2024 · Multi-objective reward generalization: improving performance of deep reinforcement learning for applications in single-asset trading · source code.
• 2019 · Spirit: Multifunctional framework for atomistic spin simulations · GitHub repository.
• Community organizer, AI Tinkerers Berlin.
• Community organizer, AI Tinkerers Cologne.

Education

• M.Sc. Physics, RWTH Aachen University. Thesis on basis-expansion approximations for functional renormalization group methods.
• 2017 · Bachelor thesis, Forschungszentrum Jülich (PGI-1) / RWTH Aachen. Modelling in-plane spin currents in atomistic spin dynamics.