Staff Software Engineer - Agentic AI Compute
Core
Building the compute platform for AI at scale, specifically orchestrating models, data, tools, and services across distributed workflows to optimize CPU performance for agentic AI systems.
Role type
Staff Software Engineer (Agentic AI Compute)
Builds
CPU performance optimization for next-generation AI data centers (Arm Neoverse, AGI-CPU) and distributed AI workloads.
Domain
Cloud computing, AI infrastructure, CPU architecture (Arm/AArch64)
Deliverable
production ML models
Required skills
C++, Python, Linux development, profiling and debugging, end-to-end system performance analysis, parallel programming, low-level optimization (SIMD/Neon/SVE/SME)
Preferred skills
ML frameworks (PyTorch, vLLM, SGLang), CPU optimization in heterogeneous systems, agentic AI frameworks, vector search/retrieval, containerization, distributed AI workloads (batching, scheduling, KV-cache), open-source contribution
Technologies
Arm Neoverse, AGI-CPU, C++, Python, PyTorch, vLLM, SGLang, llama.cpp, Linux, SIMD, Neon, SVE, SME
Responsibilities
Debug complex issues in unfamiliar software stacks, identify performance opportunities, deliver sophisticated technical work involving multiple contributors, optimize software for AArch64/Arm servers, engage with the open-source community
Seniority
Staff, hands-on IC with mentoring responsibilities
