Software Engineer III, AI Edge Applied ML Experiences

In this role you will advance multi-task generative models running directly on devices to power real-time image editing. You will build clean C++ APIs and native wrappers for Android and iOS to integrate ML experiences into partner apps like Google Photos, Snapseed, and YouTube. You’ll enable private, local intelligence with multimodal models and optimize edge pipelines for real-time performance. You’ll collaborate across teams to push on-device ML capabilities from prototyping to production at scale.

• bonus target (15%)
• equity
• benefits

• Develop on-device multi-task generative models and real-time image editing features
• Create reusable cross-platform C++ APIs and Android/iOS native wrappers for seamless integration
• Integrate and deploy multimodal on-device models (e.g., Gemma) for local data processing across text, audio, image, and video
• Optimize and port goal-segmentation and edge pipelines to mobile GPU/TPU environments for real-time tracking
• Design low-latency audio and speech pipelines (e.g., direct translation) using optimized C++ and MediaPipe graphs

• Bachelor’s degree or equivalent practical experience
• 2 years of software development experience in C++ or 1 year with advanced degree
• 1 year of experience with ML infrastructure (deployment, evaluation, data processing, debugging)
• Experience with on-device ML SDKs/tooling (MediaPipe, TensorFlow Lite, ExecuTorch, Core ML, SNPE/QNN)

• Excellent communication and collaboration skills
• Proven track record of delivering AI/ML on-device projects
• Adaptability and fast-paced execution
• C++
• MediaPipe
• TensorFlow Lite