Seiji Eicher
Distributed LLM Inference
Anyscale
Seiji is a software engineer and a maintainer of the open-source Ray Serve LLM library at Anyscale. Ray Serve LLM is an orchestration framework for distributed LLM inference with support for vLLM and SGLang, and supports advanced serving patterns, including wide-EP, prefill/decode disaggregation, and prefix aware request routing.
Seiji’s Sessions
This is where the speaker sessions are displayed
My sessions
Speaker
Don't delete!! this section is used to get the heading of the modal for gtm tracking contains logic for gtm tracking when modal open