Yichi Zhang is an SGLang developer focused on efficient LLM inference, with an interest in KV cache management and optimization. He is also a core developer of sglang-omni, building inference engines for omni-modal, TTS, and ASR models. He leads community education and developer relations efforts within the SGLang community and is the initiator and a core developer of zero-to-sglang, SGLang’s official course.