AI Performance Engineer - New Grad
Core
Bring up and optimize open-source foundation models (DeepSeek, Llama, Qwen, GPT OSS) on SambaNova's dataflow platform to improve inference throughput and latency.
Role type
New grad AI Performance Engineer (IC)
Builds
Optimized inference performance for agentic workloads on SambaNova hardware
Domain
AI Infrastructure / High-performance Computing
Deliverable
production ML models
Required skills
Python, C++, algorithms, data structures, numerical computing, PyTorch, TensorFlow, JAX, vLLM, TensorRT, SGLang, CUDA, Triton
Preferred skills
LLM training or inference, quantization, graph optimization, kernel fusion, model partitioning, memory hierarchy knowledge, scheduling for large models, ML systems research experience
Technologies
SambaNova RDU, SambaRack, DeepSeek, Llama, Qwen, GPT OSS, PyTorch, TensorFlow, JAX, vLLM, TensorRT, SGLang, CUDA, Triton
Responsibilities
Bring up new foundation models on the SambaNova platform; Profile model performance across compiler, runtime, and hardware layers; Identify bottlenecks and propose dataflow or scheduling optimizations; Apply techniques from recent research in quantization, scheduling, and memory optimization; Benchmark throughput and latency; Work with ML, compiler, runtime, and hardware teams to ship performance improvements
Seniority
Junior, hands-on IC
