Optimize Model Inference on CPUs

Almost all quality control and assurance are done manually. Whole teams review each image for potential issues, spending up to 30% of their time on this task.

AI Consensus Scoring at Hasty* finds potential issues. A user can accept or reject its suggestions. Its algorithm uses neural networks and is memory consuming so Hasty is investigating using oneAPI CPU accelerators.


Bogdan Zhurakovskyi is a machine learning engineer at Hasty.ai. He received his PhD from the National Technical University of Ukraine, Igor Sikorsky Kyiv Polytechnic Institute.​