Senior Agentic Systems & Applications Engineer
Posted Updated
Description
Advanced Micro Devices, Inc. (AMD) seeks a Senior Agentic Systems & Applications Engineer to design, deploy, and optimize intelligent AI systems for HPC and GPU workloads.
You will enable environments, orchestrate workloads, diagnose failures, and push performance improvements across applications, libraries, runtimes, and kernels. You will collaborate across platform, performance, compiler, runtime, and AI teams to translate research into robust, scalable solutions that advance AMD’s AI and HPC
Design and develop agentic AI frameworks for planning, tool use, memory, reasoning, and long-running workflows. Build automation for environment setup, containers, dependency management, and reproducible deployment. Launch scheduler-based workloads with resource-aware configurations and policies. Investigate failures via logs and telemetry and craft automated remediation strategies. Profile applications to identify performance bottlenecks and optimization opportunities. Benchmark and validate performance improvements with robust methodologies. Contribute to optimization workflows across applications, libraries, runtimes, compilers, and GPUs. Develop testing, observability, validation, rollback, and performance regression mechanisms. Explore new agentic AI approaches and translate research into practical solutions. Master’s or PhD in Computer Science, Computer Engineering, Electrical Engineering, or related technical discipline. Experience developing and deploying agentic AI or autonomous software systems (orchestration, planning, tool use, memory, long-running workflows). Experience building or integrating LLM inference and serving solutions (on-prem or cloud). Strong programming in Python with one systems language (C, C++, Rust, Go). Experience building reliable production software with CI/CD, observability, and debugging practices. Experience connecting automation or AI with build/test systems, schedulers, telemetry. Collaborative approach to solving complex problems across engineering disciplines.Skills
Sign in to read and join the discussion.