Senior software developer with 15 years of experience, including about six years specializing in Python.

Team lead and responsible for architecture at my own limited liability company (GmbH) since 2022.

Machine Learning. Focus on next-token prediction (NTP), automatic speech recognition (ASR), and computer vision. In-depth expertise in Transformer architectures and self-supervised pre-training. Experience training large models on multi-GPU systems: distributed training, throughput, and memory optimization.

Systems & Infrastructure. Development of highly available systems and the associated network and storage infrastructure; HA architectures, HAProxy, nginx, network-based storage, nftables firewalls, and policy-based routing. Reproducible deployments with Nix/NixOS and MicroVMs. Performance-critical components in Rust; author of Gradient, a CI system for NixOS.

Embedded. Development for microcontrollers (ESP32, CH32 series, Arduino), now primarily in Embedded Rust; previously in C++. Experience with sensor fusion and real-time control loops, among other things.

This provides the entire chain from a single source: model development and distributed training, the infrastructure for reliable and reproducible operation, all the way through to execution on resource-constrained hardware.

Originally from Java and C# development, later C++; today, a clear focus on Python and Rust.