Research Engineer (On-site)

Job summary

A Research Engineer is sought to advance the capabilities of an agentic AI platform for semiconductor design automation. The role involves designing and conducting experiments, developing agents, curating datasets, and creating evaluations while collaborating closely with engineering teams to implement improvements.

Responsibilities

• Develop multi-agent systems for code generation that interact with EDA tools, proposing fixes and iterating through chip design and verification processes.
• Create research prototypes that integrate with production tools and collaborate on productionizing successful outcomes.
• Establish reinforcement learning environments and evaluations for agents, exploring proxy rewards and balancing speed and accuracy of custom tools.
• Generate datasets from silicon collateral and maintain data quality and licensing.
• Conduct disciplined analyses of experiments, documenting results and driving technical progress.
• Stay updated on developments in LLM agents, reinforcement learning, and program synthesis.

Qualifications

Bonus Points

• PhD in Computer Science, AI, or ML, or equivalent research experience, with publications in relevant fields.
• Proficient in Python and ML frameworks, preferably PyTorch.
• Proven ability to translate research into functional systems.
• Experience in designing evaluation environments and reward models for sequential tasks.
• Familiarity with EDA tools and data acquisition/curation processes.
• Strong communication skills for effective collaboration with engineering teams.
• Research experience in program synthesis, constrained decoding, or execution-based rewards.
• Background in offline reinforcement learning from tool traces or human corrections.
• Contributions to open-source projects related to reinforcement learning or AI frameworks.
• Knowledge of semiconductor or complex technical domains.
• History of successfully transitioning research into production.