About the role#
The Seed Infrastructures team works on distributed training, reinforcement learning frameworks, high-performance inference, and heterogeneous hardware compilation for AI foundation models. This internship is for PhD students interested in the intersection of compiler technology and large-scale machine learning systems.
What you'll need#
- You are currently pursuing a PhD in Computer Science, Electrical Engineering, or a related technical field.
- You have experience using or developing open-source frameworks for LLM inference like vLLM or SGLang.
- You are proficient in at least one deep learning framework such as PyTorch, Megatron, DeepSpeed, or JAX.
- You have experience with model inference workflows.
- You understand modern computing systems, including how hardware, storage, and networking impact machine learning workloads.
- You are familiar with compilers or model optimization pipelines like PyTorch Dynamo.
- You have experience with distributed or large-scale machine learning systems, including FSDP, DeepSpeed, or GSPMD.
- You have experience with GPU, TPU, or NPU programming and performance optimization, or high-performance computing such as CUDA, Triton, NCCL, or RDMA.
- You have an understanding of AI compiler and model optimization stacks like torch.fx, XLA, or MLIR.
- You can commit to working for 12 weeks in 2026.
What you'll do#
- Contribute to AI compiler optimizations for both training and inference workloads.
- Develop and extend MLIR-based compiler passes for graph lowering, optimization, and code generation.
- Optimize model execution on GPU and NPU accelerators to improve performance, memory efficiency, and scalability.
- Support model deployment pipelines, including compilation, packaging, and runtime integration.
- Assist with distributed training and inference acceleration, such as parallel execution, communication optimization, and runtime scheduling.
- Benchmark, profile, and analyze the performance of large-scale models across different hardware backends.
- Collaborate with researchers and engineers to translate system requirements into compiler and runtime improvements.
Location & details#
- Location: San Jose, California
- Modality: On-site
- Duration: 12 weeks in 2026
- Employment Type: Full-time intern
- Compensation: $85 per hour


