Agentic AI Engineer for Compute & Hardware Optimization

Socket.dev; Santa Clara (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

AMD in Santa Clara, CA seeks an AI Engineer to build recursive self‑improvement systems for compute workloads. This role spans AI systems, performance engineering, and hardware‑aware optimization, turning complex tasks into repeatable loops with clear inputs and metrics.

You will collaborate with researchers and engineers to design end‑to‑end optimization, measurement, and validation pipelines that scale and improve over time.

Qualifications

• Hands-on experience with AI systems and hardware-aware optimization.
• Experience evaluating AI/ML systems with objective metrics.
• Ability to design reliable experiment loops and benchmarks.

Responsibilities

• Build AI system optimization loops for compute workloads and hardware workflows.
• Develop systems to generate, compile, test, benchmark, profile, and iterate on candidate improvements with minimal human intervention.
• Convert high‑value engineering workflows into verifiable tasks with clear graders and metrics.
• Collaborate with AI researchers on reward design, shaping, and model improvement loops.
• Design feedback systems that collect data from successful and failed attempts, profiler traces, benchmarks and logs.
• Improve iteration speed through staged validation, caching, and parallel execution.

Skills

Python

C++

CUDA

HIP

Reinforcement Learning

Performance profiling

System optimization

Collaboration

Education

Bachelor's degree in Computer Science, Electrical Engineering, or related field

Master's preferred; PhD a plus

Job description

AMD in Santa Clara, CA seeks an AI Engineer to build recursive self‑improvement systems for compute workloads. This role spans AI systems, performance engineering, and hardware‑aware optimization, turning complex tasks into repeatable loops with clear inputs and metrics.

You will collaborate with researchers and engineers to design end‑to‑end optimization, measurement, and validation pipelines that scale and improve over time.