Professional Summary
Jeff XING is an AI Product Engineer at AMD. His work focuses on LLM and multimodal inference frameworks, including SGLang, vLLM, and ATOM, for agentic and distributed inference systems.
Education
Master of Computing
2019-01-01
2020-01-01
National University of Singapore
Bachelor of Engineering (Electrical Engineering)
2012-09-01
2016-06-01
Beihang University
Interests
LLM and Multimodal Inference
Agentic AI Systems
Distributed Inference Systems
SGLang / vLLM / ATOM
ISP / AIISP
Photography Technology
3D Vision (Sensing / Reconstruction / Understanding / Generation)
📚 My work interests
My current work and broader technical interests include:
- High-performance LLM and multimodal inference frameworks
- SGLang, vLLM, and ATOM
- Inference optimization and scalable serving
- Agentic AI systems
- Efficient inference for agent workflows
- Reliable orchestration across models and tools
- Distributed inference systems
- Scaling inference across heterogeneous hardware
- Improving throughput, latency, and resource efficiency
- Computer vision and 3D vision
- World sensing, understanding, and generation
- 3D sensing, reconstruction, and simulation for robotics
- Machine vision
- ISP, AIISP, and photography technology
- AIISP architecture
- AI-enhanced ISP pipelines, calibration, tuning, and image quality Let’s build intelligent systems that sense, understand, and interact with the world! 😃
Publications & Patents
(2026).
A Light Source Estimation Method by ***.
WIPO Patent Application No. ***.
(2026).
Shading Correction Method by ***.
WIPO Patent Application No. ***.
(2026).
Tone Mapping Method for ***.
WIPO Patent Application No. ***.
(2026).
WFOV Rendering Method by ***.
WIPO Patent Application No. ***.
(2025).
*** based Autoexposure and Tone Mapping Method.
WIPO Patent Application No. ***.