SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
OpenAI is hiring a Research Engineer to apply advanced AI systems to semiconductor engineering challenges. The role sits at the intersection of AI research, model training, and hardware expertise, with the goal of helping chip engineers develop better designs and shorten development cycles.
You will own experiments end-to-end: from initial concept through implementation, analysis, and iteration. This includes building reinforcement learning environments and evaluation frameworks for chip-design tasks such as RTL generation, design verification, and physical design optimization. You'll develop and test approaches that enable AI models to use chip-design tools effectively while optimizing for power, performance, and area without sacrificing correctness.
Key responsibilities include designing rigorous experiments with clear baselines, measuring whether improvements generalize to new tasks and designs, investigating failures across model behavior, reward functions, evaluation infrastructure, and experiment systems, and improving iteration speed through better tooling and proxy rewards. You'll also translate successful experiments into reusable research code and training workflows, collaborating closely with researchers and hardware specialists.
The role emphasizes strong coding fundamentals, careful experimental judgment, and independent problem-solving. You'll stay close to implementation, explain what you built and what you learned, and communicate progress clearly across research, software, and hardware teams. Prior chip-design experience is helpful but not required—you can learn the domain alongside the team's hardware specialists.
REQUIREMENTS:
- Strong programming and debugging skills with a track record of turning technical ideas into working software
- Experience with reinforcement learning, model evaluations, post-training, or applied ML research
- Experience building tool-using agents, reward functions, or automated evaluation systems
- Ability to form clear hypotheses, design useful experiments, and distinguish meaningful results from noise or evaluation errors
- Capacity to work independently on ambiguous problems and make practical decisions about what to build or test next
- Commitment to staying close to implementation and explaining what you built, what failed, and what you learned
- Strong collaboration and communication skills across research, software, and hardware teams
- Commitment to developing safe, beneficial AI
NICE TO HAVE:
- Familiarity with experiment orchestration, distributed training, or research infrastructure
- Experience with RTL, Verilog/SystemVerilog, EDA tools, formal verification, or chip-design automation
Note: Candidates may need to meet certain legal status requirements to comply with U.S. export control laws and regulations.
