Software Engineer, Inference - Multi Modal

About this role

Join OpenAI's Inference team to build scalable infrastructure for serving multimodal AI models including image, audio, and speech capabilities in production. You'll work on high-performance systems that handle complex, heterogeneous workloads while collaborating closely with research and product teams.

What you'll do

• Design and implement inference infrastructure for large-scale multimodal models
• Optimize systems for high-throughput, low-latency delivery of image and audio inputs/outputs
• Enable research workflows to transition into reliable production services
• Collaborate with researchers, infrastructure, and product teams on model deployment
• Improve system-level performance including GPU utilization and tensor parallelism
• Own end-to-end problems in distributed compute and data handling systems

What they're looking for

• GPU-based ML workload optimization
• Large language model or multimodal inference systems
• Distributed computing and networking
• Inference frameworks (vLLM, TensorRT-LLM, or custom systems)
• High-throughput data pipeline design
• Image and audio processing systems
• Systems engineering and performance tuning
• Model parallelism and hardware abstraction