At Noetive, you'll push the frontier of what foundation models can do across language, vision, and world model development, designing and running experiments on post-training, reasoning, agentic behavior, and evaluation, then translating those findings into capabilities that ship in real products. You'll work with a small, principal-level team that moves fast from hypothesis to result, own the full arc from literature review through ablations to writeup, and hold a high bar for rigor even under startup timelines. We're looking for a PhD or equivalent research track record, deep fluency in modern deep learning (PyTorch, distributed training, LLM fine-tuning and RL), a strong publication or open-source footprint, and the instinct to pick problems that matter rather than problems that are merely publishable.