Research Engineer, Chip Design — Reinforcement Learning

Job Title

Research Engineer, Chip Design — Reinforcement Learning

Role Summary

Join the Code RL team to develop reinforcement-learning methods and infrastructure that improve models' capabilities for silicon and RTL design. The role combines research and engineering work to build RL environments, metrics, experiments, and production training pipelines used for hardware design tasks.

Experience Level

Mid-level. The posting does not state a specific years-of-experience requirement; candidates should have practical experience in chip design and RL workflows.

Responsibilities

Contribute to research and engineering efforts that apply RL to chip design and verification.

• Design and implement RL environments for RTL generation and verification.
• Optimize EDA-tool latency and develop proxy reward functions for long-running toolchains.
• Design, run, and analyze experiments to shape the research roadmap.
• Deliver research outputs into production training runs and infrastructure.
• Collaborate with researchers and engineers across teams.
• Build tooling and automation for synthesis, simulation, and verification flows.

Requirements

Core technical skills and domain experience required; nice-to-have items listed after must-haves.

• Must-have: Expertise in ASIC or FPGA design, including RTL and design verification.
• Must-have: Practical experience with industry EDA tools and chip design flows.
• Must-have: Experience applying reinforcement learning in evaluations or production settings.
• Must-have: Ability to move research ideas into engineering implementations and production runs.
• Nice-to-have: Background in ML accelerators or high-performance compute hardware.
• Nice-to-have: Familiarity with high-level synthesis, architecture simulators, or chip-design automation tools.
• Nice-to-have: Experience optimizing tool latency or designing proxy rewards for EDA workflows.

Education Requirements

Not specified.

About the Company

Company: Anthropic

Headquarters: San Francisco, CA, United States

Anthropic is an AI research and safety company developing reliable, interpretable, and steerable large language models (Claude) and related tools, focusing on alignment, reinforcement learning, and scalable infrastructure for safe AI deployment.