About the roleWe're looking for research engineers with a strong computer vision background who believe that visual and spatial reasoning are core to fully unlocking the capabilities of LLMs. In this role, youll work on research, development, and evaluation for state-of-the-art Claude models, with a focus on visual and spatial capabilities. This role is highly collaborative and will touch many aspects of our broader research efforts, taking a full-stack approach across pretraining, RL, and runtime techniques like agentic harnesses. Additionally, you'll partner with the product org to ensure that the vision improvements you deliver impact Claude's performance on real-world tasks.What youll do:Run experiments to evaluate architectural variants, data strategies, and SL and RL techniques to improve Claude's visionDevelop and test tools, skills, and agentic infrastructure that enable Claude to reason over visual inputsCreate evaluations and benchmarks that measure progress on multimodal capabilities across training and deploymentWork with our product org to find solutions to our most vexing API customer challenges related to vision and spatial reasoningYou may be a good fit if you:Have 7+ years of ML, computer vision, and software engineering experience through industry, academia, or other projectsAre familiar with the architecture, training, and operation of large vision language modelsHave experience creating and evaluating large synthetic and real-world visual training datasetsHave experience engaging in systematic prompting, finetuning, or evaluationAre results-oriented, with a bias towards flexibility and impactEnjoy pair programming and cross-team collaborationCare about the societal impacts of your workStrong candidates may also have experience with:Large-scale pretraining, SL, and RL on language modelsDeep learning research on images, video, or other modalitiesDeveloping complex agentic systems using LLMsHigh-performance ML systems (GPUs, TPUs, JAX, PyTorch)Large-scale ETL and data pipeline developmentRepresentative projects:Running experiments to determine ideal training datamixes and parameters for a synthetically generated vision datasetFinetuning Claude to maximize its performance using a particular set of agent tools/skillsBuilding a pipeline to ingest and process a novel source of visual training dataDesigning and running experiments to evaluate the scalability of two architectural variants