AI has achieved significant progress in recent years, with models achieving human-level performance in numerous tasks. However, the real challenge lies not just in training these models, but in implementing them optimally in everyday use cases. This is where AI inference becomes crucial, arising as a primary concern for experts and industry profess