Research Engineer, Generative Video

About this role

Mirage is hiring a Research Engineer to develop and optimize large-scale video generation models, focusing on making advanced generative systems faster, more efficient, and production-ready. You'll work at the intersection of deep learning research and systems engineering, tackling novel modeling approaches, training infrastructure, and real-time inference optimization.

What you'll do

• Train and optimize large-scale video and multimodal generative models
• Improve efficiency across training and inference through memory, latency, and cost optimization
• Implement model acceleration techniques including distillation, quantization, and pruning for diffusion and autoregressive generation
• Build and maintain distributed training systems with GPU optimization and parallelism
• Develop experimentation, evaluation, and debugging tools for model development
• Translate research prototypes into production-ready systems and monitor real-world performance

What they're looking for

• Deep learning systems and infrastructure
• PyTorch and CUDA programming
• Triton kernel optimization
• Distributed training frameworks (FSDP, etc.)
• Model optimization techniques (quantization, pruning, distillation)
• GPU performance profiling and debugging
• Low-latency inference optimization
• Prototype-to-production development velocity

Benefits

• Medical, dental, and vision coverage
• 401(k) with employer match
• Commuter benefits and catered meals multiple days per week
• Dinner stipend for late-night work and Grubhub subscription
• Health and wellness perks plus multiple team offsites and monthly events
• Generous paid time off policy