Introduction

I possess over three decades of experience in software engineering and have a passion for creating solutions, iterating from design to deployment until success, taking responsibility and often leading the effort. I am at home in computer graphics & multimedia, applied math, cryptography, driver development, algorithms and high performance solutions, from complete POSIX systems to bare-metal embedded devices.

I enjoyed employment at Sun Microsystems ($\rightarrow$Oracle) and ATI ($\rightarrow$AMD) at their headquarters, thereafter I continued as a contractor. I served a broad range of companies, such as Zafena, Mathworks, Google, Harman/Becker, Siemens AG Munich and IBM Frankfurt.

Early 2020 I picked up contemporary C++ again for new challenges, covering C++11 to C++23. I utilize cmake, gcc, clang/llvm with sanitizer, valgrind, clang-format, clang-tidy, Catch2, Jenkins, GitLab beside others for code validation, performance analysis and CI/CD. Some of these projects are:

• Gamp, replicating JogAmp's experience of GraphUI and overall platform agnostic graphics framework for native platforms using C++20 and for web browser via WebAssembly using emscripten. This work is also based on gfxbox2, evaluating C++20 and the web-target.

• Direct-BT, is a high-performance Bluetooth LE and BREDR host library hooking up to Linux's kernel BlueZ implementation, circumventing D-Bus and supporting secure client & server modes. It has been used successfully in a medical trial, as well as in a connected medical device terminal. Implemented in C++17/20 to be used directly from native code or via a thin Java layer.
  Presentations: Latest Blogs, BLE Programming with C++ & Java

• Cipherpack provides encrypted & authenticated (OTA) streaming and atomic file operations used to implement a secure system and application upgrade across a wide range of media for devices in the medical space. Implemented in C++17/20 to be used directly from native code or via a thin Java layer.

I maintained JogAmp for over a decade, a cross-platform Audio/Video, 3D Graphics and Processing library for Windows, Linux, MacOS, iOS, Android and bare metal embedded devices. While it runs on the Java VM, the implementation layers involve our own compiler, C/C++, ObjectiveC native system level code. Tasks included low level debugging and performance analysis as well as coding computer graphics and adding a video layer with FFmpeg. This culminated to our own resolution-independent GPU Curve Rendering GraphUI toolkit. Jenkins & JUnit across the target network covered our CI/CD tasks.
Presentations: Latest Blogs, JavaOne 2008, Tegra-1 Video Demo, Siggraph[2010, 2011, 2012, 2013, 2014], Fosdem[2013, 2014].

Most of the above projects are open sourced and available on jausoft and github.

I run a little computer-science class, covering C++ basics, gfxbox2, hanoi, pacman.

I just started hacking Rust and some AI projects sparked my interest in NN again.

The following sub-section offers some more excerpts from my background, if interested.