Senior Software Engineer, On-Device Machine Learning
Sunnyvale, CA, USA · Senior · Full-time · Posted Oct 29, 2026
- Experience
- 5 years software development; 3 years testing, maintaining, or launching software products and 1 year software design and architecture; 3 years in speech/audio, reinforcement learning, ML infrastructure, or another ML specialization; 3 years end-to-end Machine Learning. Preferred: 5 years with data structures and algorithms.
- Education
- Bachelor’s degree or equivalent practical experience.
About the job
Develop and optimize LiteRT, Google’s on-device AI framework, to deploy and accelerate machine learning models across edge devices and platforms.
What you'll do
- Develop LiteRT and enable hardware-accelerated use cases on edge platforms.
- Enable deployment of models such as Gemini Nano and Gemma across accelerators and platforms including Android and Chrome.
- Optimize on-device inference through model conversion, runtime, and kernel improvements; write, test, and troubleshoot software.
What they ask for
- Bachelor’s degree or equivalent practical experience.
- 5 years of software development experience in one or more programming languages.
- 3 years testing, maintaining, or launching software products, plus 1 year in software design and architecture.
- 3 years in a listed ML specialization and 3 years of end-to-end Machine Learning experience.
Nice to have
- Master's degree or PhD in Computer Science or a related technical field; 5 years of experience with data structures and algorithms.
- Experience with on-device ML SDKs/tooling, ML converters/compilers, hardware-accelerated inference, ML frameworks, or generative AI model optimization.
Pay
$174000 - $252000 USD base salary
Tools
LiteRT, TensorFlow Lite, Gemini Nano, Gemma, GPU, Pixel TPU, NPUs, CPU
AI focus
Build and optimize on-device machine learning infrastructure and inference, including deployment of generative AI models across edge devices and hardware accelerators.
Model-written summary · the original posting is authoritative