Member of Engineering (Multimodality - Research Lead)

Build a new multimodality team from the ground up, teaching a frontier coding model to “see” through image input. Own the technical direction from adapter-based approaches toward native multimodality, shipping first image-input capabilities. Partner across Applied Research and collaborate with evals, architecture, data, and post-training teams to create datasets, run end-to-end experiments at scale, and land multimodal SWE capabilities.