Generative AI - ML System Engineering

About the Company

Meshy is a market leader in 3D generative AI, providing a full pipeline for 3D content creation ranging from text/image to 3D, texturing, and animation. Trusted by enterprises like Meta, Square Enix, and Deepmind, Meshy aims to make 3D creation boundless and accessible.

Responsibilities

• Co-design the next frontier of 3D & Spatial AI alongside researchers
• Build and debug training infrastructure using PyTorch for maximum parallelism and efficiency
• Optimize distributed model training across hundreds to thousands of GPUs
• Implement and maintain 3D specific custom operators in Triton or CUDA
• Develop and maintain novel data-loading frameworks and libraries
• Build efficient inference endpoints for complex multi-stage model pipelines
• Optimize models through compilation, fusion, and quantization

Requirements

• Experience in machine learning or high performance graphics
• Solid practical understanding of PyTorch or JAX
• Strong ability to write maintainable code in Python and/or C++
• Performance and efficiency oriented mindset
• Strong communication skills for a globally distributed team

Preferred Qualifications

• Hands-on experience with PyTorch internals (torch.compile, FSDP2)
• Experience building Triton kernels
• Experience with large-scale distributed training (DP, TP, CP, PP, ZeRO)
• Experience with diffusion models in 3D or video
• Experience with low precision bf16 or fp8 training