Senior Embedded ML Software Architect - Mobile

Qualcomm; San Diego (CA); On-site; USD 235,000 - 352,000; Full time

Job summary

Qualcomm Technologies, Inc. seeks an experienced Embedded ML SW Architect to advance Snapdragon Mobile AI software. You will analyze requirements, design high-level architectures for new use-cases, and push performance/power improvements across SW/HW stacks.

You will collaborate with software/hardware teams, customers, and partners to enable on-device LLM/SLM workloads and context-driven agentic experiences within tight power and memory budgets.

Qualifications

• Experience with AI frameworks and on-device deployment of ML models.
• Strong background in mobile AI software architecture and performance analysis.
• Experience with resource-constrained embedded devices and heterogeneous compute.

Responsibilities

• Develop architecture for AI features on Snapdragon Mobile products.
• Research requirements and prototype across CPU, GPU, NPU workloads.
• Architect on-device agentic experiences with context/memory management.
• Design telemetry-driven scheduling across heterogenous compute with live signals.
• Manage memory/state lifecycle for on-device LLM/SLM serving.
• Collaborate with partners to align on roadmap and milestones.
• Monitor AI trends and contribute to strategic direction of product.

Skills

AI architectures

Embedded systems

C/C++ programming

Python programming

Performance optimization

Education

Bachelor’s degree in Computer Science, Engineering, Information Systems, or related field

Master’s degree in Computer Science, Engineering, Information Systems, or related field

PhD in Computer Science, Engineering, Information Systems, or related field

Tools

TensorFlow

PyTorch

Job description

Qualcomm Technologies, Inc. seeks an experienced Embedded ML SW Architect to advance Snapdragon Mobile AI software. You will analyze requirements, design high-level architectures for new use-cases, and push performance/power improvements across SW/HW stacks.

You will collaborate with software/hardware teams, customers, and partners to enable on-device LLM/SLM workloads and context-driven agentic experiences within tight power and memory budgets.