AI Research Engineer

Normal Computing is seeking a full-time AI Research Engineer to advance their agentic code generation tool through innovative research in large language models (LLMs) and reinforcement learning (RL). The role involves designing and conducting experiments, developing agents, and curating datasets from complex technical documents, such as chip specifications. The engineer will also create evaluation frameworks, analyze experimental results, and collaborate with engineering teams to implement improvements. Candidates should hold a PhD in computer science, artificial intelligence, or machine learning, or possess equivalent research experience, ideally with publications in relevant fields like multi-agent RL or agentic AI. Proficiency in Python and familiarity with machine learning frameworks, particularly PyTorch, are essential, along with a strong focus on reproducibility in research. This remote position is suitable for individuals who are passionate about turning research into practical applications and have a solid understanding of evaluation metrics for sequential tasks.