Senior Windows AI/ML Platform Architect

MaC VC; San Diego (CA); On-site; USD 224,000 - 335,000; Full time

Job summary

Qualcomm Technologies, Inc. seeks an experienced Windows platform architect to drive AI system performance and power improvements across SW/HW stacks and tools on the Snapdragon platform.

You will shape end-to-end AI delivery, ensure developer experience, and enable on-device enterprise AI workflows for Windows PCs. You will collaborate with software and hardware architects, PMs, OEMs, OS partners, and developers to optimize AI performance with GPU/NPU, while meeting security requirements.

Qualifications

• Extensive experience in architecture aware AI model system performance optimization on Windows PC/Laptop.
• Experience with on-device AI platform components and enterprise AI workflows.
• Strong leadership and communication skills with cross-site collaboration.

Responsibilities

• Understand trends in ML network design and their impact on SW/HW design.
• Analyze bottlenecks in end-to-end ML/AI workloads on Qualcomm stacks via simulation and on-device characterization.
• On-device correlation and tuning of algorithms versus pre-silicon predictions.
• Interface with cross-site teams to deliver performant reference implementations and tooling.
• Propose features for next-gen SoCs to reduce performance bottlenecks in workflows.

Skills

AI frameworks

GPU programming

NPU programming

Python

C++

Performance profiling

Architecture knowledge

Leadership

Education

Masters or PhD in CS or EE

Bachelor's in CS/engineering

Tools

TensorFlow

PyTorch

Job description

Qualcomm Technologies, Inc. seeks an experienced Windows platform architect to drive AI system performance and power improvements across SW/HW stacks and tools on the Snapdragon platform.

You will shape end-to-end AI delivery, ensure developer experience, and enable on-device enterprise AI workflows for Windows PCs. You will collaborate with software and hardware architects, PMs, OEMs, OS partners, and developers to optimize AI performance with GPU/NPU, while meeting security requirements.