Yaroslav Torziuk
Architect at Capgemini Engineering
Ukraine
Summary
🤩
Rockstar🎓
Top SchoolYaroslav Torziuk is an architect and technical lead with over a decade of C++ and systems engineering experience and five years in senior leadership roles, currently based in Ukraine. He has progressed from hands-on development and team leadership to shaping architecture at Capgemini Engineering, bringing practical delivery experience from companies like Lohika and Hewlett-Packard. His contributions to the high-profile OpenVINO open-source toolkit show deep expertise in GPU kernel optimization for ML inference (DeformableConvolution, ROIAlign, Softmax, Interpolate and more), highlighting a focus on performance-critical AI workloads on Intel GPUs. Yaroslav combines low-level performance tuning with system- and team-level thinking, enabling reliable deployment of complex inference pipelines. He holds an MS in Computer Science from Odessa State Academy of Refrigeration and a long-standing track record of turning algorithmic work into production-ready solutions. Colleagues would note his rare mix of kernel-level mastery and pragmatic architecture leadership.
5 years of coding experience
10 years of employment as a software developer
Master of Science (MS), Computer science, Master of Science (MS), Computer science at Odessa State Academy of RefrigerationBachelor's degree, Computer Science, Bachelor's degree, Computer Science at State Industrial Automation College, Odessa
English, Russian, Ukrainian
Github Skills (9)
kernel10
deep-learning10
gpgpu10
inference10
gpu10
openvino10
optimization10
ai9
computer-vision9
Programming languages (1)
Github contributions (5)
openvinotoolkit/openvino
Dec 2021 - Jul 2022
OpenVINO™ is an open source toolkit for optimizing and deploying AI inference
Role in this project:
ML Engineer
ML EngineerContributions:86 reviews, 4 commits, 15 PRs in 7 months
Contributions summary:Yaroslav contributed to the OpenVINO toolkit by modifying the clDNN and kernel selector code. Their work involved implementing and optimizing GPU kernels for various operations, specifically focusing on DeformableConvolution, ROIAlign, Slice, Assign, ReadValue, Softmax, and Interpolate. These changes aimed to enhance the performance and functionality of the toolkit, indicating a focus on improving the efficiency of deep learning inference on Intel GPUs.
inference-enginepytorchmodel-optimizerdeep-learninggpu
ytorzuk-altran/openvino
Nov 2021 - Dec 2023
OpenVINO™ Toolkit repository
Contributions:119 pushes, 18 branches in 2 years 1 month
pytorchdeep-learninggpuopenvino-toolkitcomputer-vision
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.