Video AI Engineer: Diffusion & Model Scaling

Shopee; Singapore; On-site; SGD 90,000 - 150,000; Full time

Job summary

Shopee is seeking a research-focused candidate to contribute to cutting-edge AIGC video generation at our Singapore office. The role involves designing pipelines, evaluating alignment strategies, and collaborating with engineering teams to translate research into scalable systems.

The ideal candidate will have a strong publication record, familiarity with diffusion-based models, and the ability to ship working prototypes from research ideas.

Qualifications

• Master's degree or above in Computer Science, AI, or related field; expected to graduate by May 2027 and join by June 2027.
• Strong publication record at top-tier venues.
• Familiarity with video generation architectures and diffusion-based generative models.
• Demonstrated ability to ship research into working systems.
• Excellent communication and collaboration across research and engineering teams.

Responsibilities

• Design and execute training pipelines for large-scale AIGC models, including data curation and stability optimization.
• Contribute to architecture design of video generation models from foundational training to post-training alignment and quality improvements.
• Build evaluation pipelines and explore alignment strategies to improve model output quality.
• Prototype emerging capabilities such as long-form video generation and multi-modal understanding.
• Track frontier AIGC research and deliver production-grade implementations.

Skills

Strong communication

Cross-functional collaboration

Research-to-product translation

Education

Master's degree or above in Computer Science / AI or related field

Tools

PyTorch

TensorFlow

Diffusion models

Job description

Shopee is seeking a research-focused candidate to contribute to cutting-edge AIGC video generation at our Singapore office. The role involves designing pipelines, evaluating alignment strategies, and collaborating with engineering teams to translate research into scalable systems.

The ideal candidate will have a strong publication record, familiarity with diffusion-based models, and the ability to ship working prototypes from research ideas.