Bio
Yuchen Fama is a Builder, Benchmarker, and Senior Principal Product Manager of Inference at Red Hat and also a contributor to vLLM and GuideLLM. She has more than 15 years of experience in ML and AI. She has served as VP of Product, CTO, and CPO at multiple AI startups and previously led AI/ML research teams within several Fortune 500 companies. She holds a Ph.D. in Statistics and enjoys reading, traveling, skiing, and scuba diving.
Session (1)
Additional Speaker Details
Using publicly available information we constructed an analysis to help you get a feel for this speaker before deciding to attend their session.
Yuchen Fama is a Red Hat software engineer working hands-on with vLLM and llm-d, co-authoring technical articles on distributed inference architecture, five-dimensional parallelism (tensor, pipeline, expert, data, context), prefill/decode disaggregation, and production deployment of LLM serving stacks. Her personal GitHub hosts config-optimizer and tuning tooling for vLLM + llm-d on Kubernetes/OpenShift. Attendees wanting a practitioner's view of how Red Hat's inference platform is built and deployed at scale will benefit from her session.
Speaking style
GitHub
@cemigo114- cemigo114/intelliconfig★ 0Lightweight LLM serving configuration optimizer for vLLM + llm-d
- cemigo114/claude-skills-RH-llm-d★ 1Claude skills for llm-d by Red Hat — covers OpenShift deployment, performance diagnosis, and RoCE networking
- cemigo114/llm-d-tuning-mooncake★ 1LLM-d tuning experiments with Mooncake
- cemigo114/easy-Kueue-UI★ 0Easy UI on top of Kueue for GPU resource orchestration
- cemigo114/llm-d-vllm-doc-mvp★ 0MVP demo of a use-case-oriented doc site for llm-d and vLLM