You will own projects from initial research questions through working systems and demonstrated impact, delivering advances in models and agents that solve increasingly complex technical problems, generalize across projects and domains, and measurably improve the quality and productivity of software and hardware development. A strong publication record in language models, AI agents, reinforcement learning, code intelligence, automated reasoning, or related areas, including venues such as NeurIPS, ICML, ICLR, ACL, and EMNLP.