Jimmy Gou

About Jimmy Gou

I hold a Ph.D. in Machine Learning from Tsinghua University, specializing in image recognition. My background spans machine learning and digital signal processing, including earlier work on video codecs before returning to focus on ML.Currently, I serve as a Principal Architect specializing in LLM inference optimization for edge and on-device development. My role involves hands-on prototyping and leading a team to compress and adapt pre-trained LLM models for efficient deployment on resource-constrained devices. We employ a variety of advanced techniques such as quantization, pruning, low-rank decomposition, distillation, and kernel fusion to meet demanding performance requirements.Prior to this, I founded and led aiQo.ai, an AI-driven language learning startup that developed personalized solutions using deep learning and linguistics. Although the company closed in 2019 due to funding constraints, it provided valuable experience in building and deploying end-to-end AI products across speech, NLP, and adaptive learning.At LG’s San Jose Lab, I managed a team with direct hands-on technical contribution focused on AI/ML algorithms and software development, creating efficient models for computer vision, speech, and IoT/wearable devices, while collaborating closely on the convergence of AI and cybersecurity technologies.Beyond my architectural and leadership responsibilities, I maintain a passion for creative programming and enjoy prototyping with assisted coding tools. Python is my primary programming language, though I frequently experiment with others to explore new ideas.Outside of work, I am an avid language learner. In addition to English and my native Mandarin, I am conversational in Korean, Japanese, and Spanish, and often embrace new languages when traveling to broaden my cultural and cognitive horizons.

Experience

1. Principal Architect Llm Optimization and Deployment

   LG Electronics

   Dec 2020 — Present

   Lead design and implementation of LLM inference optimization on edge and on-device platformsPrototype and guide team hands-on to compress and adapt pre-trained models for deploymentApply compression techniques including quantization, pruning, low-rank, distillation, and fusionCollaborate across global teams to meet real-world performance and latency requirements

Education

• Tsinghua University, China

  M.S. Electrical Engineering

• Tsinghua University, China

  Ph.D. Electrical Engineering

• University of Electronic Science and Technology, China

  B.S. Electrical Engineering

Skills

• Mpeg
• Digital Signal Processing
• C
• Computer Vision
• Matlab
• Arm
• Mpeg2
• Parallel Computation and Implementation
• Rtos
• C++
• Algorithms
• Python
• Firmware
• Perl
• Embedded Systems
• Computation with Privacy Preserving and Security
• Linux
• Machine Learning
• Device Drivers
• Video Processing
• Video Coding
• Algorithm Design and System Architecture
• Verilog
• Soc
• Tcl
• Signal Processing
• Pattern Recognition
• H.264
• Fpga
• Embedded Software
• Iot Systems and Applications
• Digital Signal Processors
• Mpeg-4
• Computer Vision and Signal Processing
• Machine Learning Algorithm Design and Software Implementation
• Image Processing