AI Engineer, Recursive Self-Improvement for Compute
Core
Build recursive self-improvement systems for compute workloads and hardware engineering workflows where AI proposes, verifies, and iterates on improvements.
Role type
Senior IC AI Engineer (systems & hardware optimization)
Builds
Agentic optimization loops, automated validation harnesses, and experiment platforms for compute platforms
Domain
Semiconductor hardware engineering + AI systems
Deliverable
production ML models
Required skills
Python, C++, C, CUDA, ROCm/HIP, reinforcement learning, reward modeling, automated program optimization, hardware-aware optimization, compiler optimization, profiling, benchmarking
Preferred skills
GPU kernel development, distributed training/inference, formal verification, hardware/software co-design, experiment tracking, open-source contributions
Technologies
PyTorch, JAX, TensorFlow, Triton, ROCm, CUDA, HIP
Responsibilities
Build agentic and learning-driven optimization loops; Develop systems to generate, compile, test, and benchmark candidate improvements; Design feedback systems accumulating data from traces and logs; Collaborate on reward design and long-horizon optimization; Improve iteration speed via staged validation and parallel execution; Mentor engineers on self-improving AI systems
Seniority
Senior, hands-on IC
