Edge AI Engineer: Vision & Multimodal ML on Devices

Nutanix; Markham; On-site; CAD 105,000 - 155,000; Full time

Benefits offered by this job

RSU grants

Bonus program

Benefits package

Job summary

Qualcomm Canada ULC is seeking a Machine Learning Engineer to join the AI Processor Solutions team. You will help develop ML applications for Snapdragon platforms, spanning computer vision, multimodal AI, and agentic AI, with a focus on accuracy, efficiency and real-world performance on Snapdragon-powered devices.

You will collaborate with hardware and software teams to deploy AI solutions, optimize models for edge devices, and contribute to reference implementations across mobile, edge, XR and

Qualifications

• Bachelor's degree in CS, Engineering, Information Systems or related field.
• 0–2 years of relevant work or publications in ML/AI domains.
• Experience with ML frameworks (PyTorch, TensorFlow) and edge deployment.

Responsibilities

• Stay current with advances in AI/ML and foundation models.
• Research, develop and optimize AI algorithms for Snapdragon platforms.
• Design and build reference applications showcasing Snapdragon AI on mobile, edge, XR and IoT devices.
• Evaluate hardware/software trade-offs to balance performance and power.

Skills

Machine Learning

Computer Vision

Multimodal AI

Agentic AI

Python

C++

PyTorch

TensorFlow

Embedded Development

Education

Bachelor's degree in CS/Engineering

Master's or PhD (preferred)

Tools

PyTorch

TensorFlow

Python

C++

Embedded OS concepts

Job description

Qualcomm Canada ULC is seeking a Machine Learning Engineer to join the AI Processor Solutions team. You will help develop ML applications for Snapdragon platforms, spanning computer vision, multimodal AI, and agentic AI, with a focus on accuracy, efficiency and real-world performance on Snapdragon-powered devices.

You will collaborate with hardware and software teams to deploy AI solutions, optimize models for edge devices, and contribute to reference implementations across mobile, edge, XR and