Research Scientist (AI)

TL;DR

Research Scientist (AI): Advancing frontier multi-modal generation models by experimenting with cutting-edge techniques and improving model performance for real-world applications. Focus on creating novel generative architectures, training methodologies, and synthetic data pipelines to revolutionize productivity software.

Location: Must be based in Beijing, China

Company

hirify.global is a global visual communication platform empowering everyone to design and publish anything.

What you will do

• Determine and lead research initiatives aligned with frontier multimodal AI trends and product goals.
• Design and run experiments to validate hypotheses using high-end GPU infrastructure.
• Interpret experimental results and translate them into actionable insights for model improvement.
• Create novel generative model architectures and training methodologies.
• Develop advanced RL algorithms and synthetic data generation pipelines.
• Collaborate with cross-functional teams and share knowledge through documentation and research publications.

Requirements

• Must be based in Beijing
• Rich experience in developing and training foundational large-scale text-to-image/video/3D generation models.
• Strong academic track record with peer-reviewed publications in top-tier conferences like CVPR, ICCV, NeurIPS, or ICML.
• Proven experience in large-scale parallel model pre-training across hundreds of GPUs.
• Proficiency in Python, PyTorch, Diffusers, Transformers, Megatron, and DeepSpeed.
• Ability to effectively document research and collaborate across teams.