Artem Ivanov

• Developed an ASIL B component for the L3 autonomous driving system in collaboration with BMW • Optimized the component's cycle time from 8ms to 0.2ms and memory consumption from 50MB to 200KB by refactoring legacy architecture with new algorithms and data structures based on C++ template metaprogramming • Improved code quality by reducing more than 50k line repetitions by creating and maintaining an ASIL B verified code generator with Bazel, Python and Jinja

• Optimized firefighters’ small-sized radar performance by 600% (from 6s to 0.9s per image) by moving all image processing calculations on a GPU (CUDA) • Reduced amount of noise in radio calling system by 20% by creating a vocoder component • Minimized dependencies between software and hardware components and decreased product delivery time by ≈15% by implementing a bootloader library • Improved user experience for the radar project by implementing software with a plugin system to interact with the radar

• Increased accuracy of a neural network which controls mechatronic prostheses by 30% by developing a motion capture feedback system • Published 4 technical articles about the system in the IEEE journal • Spoke about the system at 2 IEEE conferences