Content Understanding Multimodal Model Algorithm Engineer-Global E-commerce
Singapore
Workplace: OnsiteFull timeFunction: Content & Editorial (Writing/Editing)Education: phdSkills: ["Communication","Collaboration"]Build multimodal large language models for intelligent e-commerce content moderation. You’ll research modality fusion across text, audio, images, video, and live-streaming, improve few-shot/zero-shot multilingual and long-sequence performance, and develop adversarial defense for AIGC content. You will also enable agent capabilities with RAG tool invocation, auto-planning, and interpretable COT generation to support autonomous, high-accuracy review decisions.

