Member of Technical Staff, Inference

Reactor is a team of industry veterans building the missing layer for world models, with experience across AI, gaming, infrastructure, and robotics. The Member of Technical Staff, Inference will optimize generative media model performance, design an in-house inference runtime, and develop low-latency, high-throughput inference systems. The role will also involve model optimization, performance profiling, and collaboration with model partner teams.

Responsibilities:

• Drive our frontier position on model performance for diffusion models
• Design and implement a high-performance in-house inference runtime
• Implement optimizations using torch.compile, custom CUDA kernels, and specialized inference frameworks
• Optimize neural network models through quantization, pruning, and architectural modifications
• Profile and benchmark model performance to identify computational bottlenecks
• Collaborate directly with model partner teams to integrate their models into our platform