AI Inference & Video Compression Engineer

Beijing Foreign Enterprise Management Consultants Co.,Ltd.; Singapore; On-site; SGD 120,000 - 180,000; Full time

Benefits offered by this job

Health insurance

Job summary

Huawei is seeking an AI Inference & Compression Engineer in Singapore to advance LLM inference acceleration, develop advanced video codecs, and optimize AI-based media coding. The role includes deploying optimized models and integrating compression techniques into deployment frameworks such as vLLM.

The candidate will work on performance and quality evaluation, including PSNR, VMAF, perplexity, latency, and throughput, ensuring robust, scalable AI/video systems.

Qualifications

• Master’s or PhD in Computer Science, Electronics, Mathematics, or related fields (PhD preferred).
• Solid understanding of video coding fundamentals incl. prediction, transform coding, quantization, entropy coding (H.265/HEVC, AV1, H.266/VVC).
• Strong understanding of Transformer architectures and attention mechanisms, memory bandwidth constraints in AI inference.
• Proficiency in Python and C/C++, with hands-on experience in PyTorch, TensorFlow, etc.

Responsibilities

• LLM Inference Acceleration: research and develop compression algorithms to accelerate LLM serving, KV cache optimization, decoding bottlenecks.
• Classical Codec Development: improve RD performance, optimize entropy coding, enhance quantization for real-world apps.
• AI-Based Media Coding: develop AI-based loop filters, optical flow, and rate control.
• Model Deployment & Fusion: bridge AI research and production, optimize models for inference and integrate compression into frameworks like vLLM.
• Performance & Quality Evaluation: assess PSNR, VMAF, perplexity, latency, and throughput for AI/video systems.

Skills

Python

C/C++

PyTorch

TensorFlow

Transformer

Memory bandwidth

H265/HEVC

AV1

H266/VVC

Education

Master's or PhD in CS/EE/Math

Tools

vLLM

Job description

Huawei is seeking an AI Inference & Compression Engineer in Singapore to advance LLM inference acceleration, develop advanced video codecs, and optimize AI-based media coding. The role includes deploying optimized models and integrating compression techniques into deployment frameworks such as vLLM.

The candidate will work on performance and quality evaluation, including PSNR, VMAF, perplexity, latency, and throughput, ensuring robust, scalable AI/video systems.