Open access
2026
End-to-End Latency Decomposition in AI-Driven Web Applications: Rethinking Infrastructure in LLM Based Systems
An end-to-end latency decomposition framework for analyzing performance in AI-powered web applications and provides a practical methodology for identifying performance bottlenecks and offers actionable insights for optimizing AI-driven web systems is introduced.
J. Hamzabegović, Amel Džanić, Kenan Duraković
· IJEEC - INTERNATIONAL JOURNA... · 0 citations