Key job responsibilities
You'll research, develop, and deploy the next generation of on-device AI for communication, bringing small language models, agentic capabilities, and generative intelligence to the edge so that customers get responsive, and private experiences without relying on the cloud. In practice, that means designing on-device arbitration systems that intelligently route between local and cloud inference in real time, applying modern compression techniques like quantization, distillation, and low-rank adaptation to fit frontier-class capabilities into tight memory and latency budgets, and optimizing inference across on-device accelerators such as NPUs and mobile GPUs. You'll build privacy-first personalization that learns and adapts without user data ever leaving the device, and contribute to the science of edge communication intelligence, intent understanding, contact ranking, proactive connectivity, and conversational reasoning, all running locally. Throughout, you'll collaborate closely with hardware, systems, and product teams to co-design models with the target device in mind, and translate research advances into production features that push the state of the art in on-device AI.
About the team
We're building up the Alexa Connections science team, so you'll join early and grow alongside it. You'll work closely with experienced scientists and engineers invested in your development, with plenty of mentorship, room to broaden your skills, and the chance to own meaningful problems as you grow. It's a supportive place to do impactful work early in your career, shaping how millions of customers connect every day.
BASIC QUALIFICATIONS
- Master's degree or above in computer science, electrical engineering, or related field
- Knowledge of programming languages such as C/C++, Python, Java or Perl
- 2+ years of experience in a university, industry, or government lab with emphasis on machine learning, deep learning, speech/NLU, or on-device AI
- Experience with ML frameworks such as PyTorch or TensorFlow
- Experience developing and deploying ML models in at least one of the following: on-device inference, model optimization, speech/audio, NLU, or LLMs
PREFERRED QUALIFICATIONS
- PhD in computer science, machine learning, engineering, or related fields
- Experience with on-device model optimization: quantization, distillation, pruning, or low-rank adaptation for resource-constrained deployment
- Experience with on-device inference runtimes (ONNX Runtime, TensorFlow Lite, ExecuTorch) or hardware accelerators (NPUs, DSPs, mobile GPUs)
- Experience building or deploying small language models (SLMs) for edge environments
The base salary range for this position is listed below. As a total compensation company, Amazon's package may include other elements such as sign-on payments and restricted stock units (RSUs). Final compensation will be determined based on factors including experience, qualifications, and location. Amazon offers comprehensive benefits including health insurance (medical, dental, vision, prescription, basic life & AD&D insurance), Registered Retirement Savings Plan (RRSP), Deferred Profit Sharing Plan (DPSP), paid time off, and other resources to improve health and well-being. We thank all applicants for their interest, however only those interviewed will be advised as to hiring status.
CAN, BC, Vancouver - 149,300.00 - 249,300.00 CAD annually
CAN, ON, Toronto - 149,300.00 - 249,300.00 CAD annually
