Software Engineer, On-Device Machine Learning

Overview In this role you will help advance on-device AI by developing and optimizing LiteRT, Google’s on-device AI framework. You will work with cross-functional teams to enable model deployment and acceleration across edge devices and platforms. You will tackle performance, runtime, and kernel improvements to push efficient ML at scale. This is a chance to shape portable AI infrastructure that powers Google products and third-party solutions. Compensation / Benefits15% bonus targetequitybenefits ResponsibilitiesCollaborate through design and code reviews to uphold best practices across available technologiesImplement solutions in ML areas and contribute to model optimization and data processingDevelop LiteRT for on-device AI, focusing on hardware acceleration and cross-platform supportEnable on-device deployment of key models (e.g., Gemini Nano, Gemma) across accelerators and platformsImprove on-device runtime and kernel performance to boost inference efficiency Key requirementsBachelor’s degree or equivalent practical experience2 years of software development experience or 1 year with advanced degree2 years of ML infrastructure experience (deployment, evaluation, optimization, data processing, debugging)Experience with runtimes and performance tuningExperience in mobile developmentcollaborationleadership qualitiesenthusiasm for solving novel problemsML frameworks: PyTorch, JAX, TensorFlowon-device ML SDKs/tooling: TensorFlow Lite, ExecuTorch, Core ML, SNPE/QNNruntime optimization