Summary
Matevž Fabjančič is a software developer with a decade of experience, currently contributing at Cosylab from Ljubljana, Slovenia. His academic work centers on compiler-level approximate computing for convolutional neural networks, aiming to make complex models feasible on power-constrained mobile devices. That research background gives him a strong blend of systems-level thinking and machine-learning pragmatism, with hands-on expertise in bringing algorithmic ideas closer to production constraints. He has roots in academia as a peer tutor at the University of Ljubljana and holds a Master's focused on applied compiler and mobile ML techniques. Colleagues can expect a pragmatic engineer who thinks about performance, energy trade-offs, and compiler optimizations beyond typical application code.
10 years of coding experience
Master's degree, MS Thesis: Compiler-Level Approximate Mobile Computing, Master's degree, MS Thesis: Compiler-Level Approximate Mobile Computing at University of Ljubljana, Faculty of Computer and Information Science
Škofijska klasična gimnazija