Embedded ML SW Architect -- Snapdragon Mobile (Principal or Director level)

Originally posted 18 October 2026 by the employer.

About the role

This role is responsible for Mobile Handset software architecture and software system design focused on Snapdragon Mobile products.

What you'll do

- Architect on-device agentic experiences, context and memory management, perception pipelines, and heterogeneous compute placement for LLM/SLM workloads across CPU, GPU, and NPU under tight power, thermal, and memory budgets. - Architect closed-loop, telemetry-driven resource scheduling for heterogeneous compute to continuously re-evaluate and re-place workloads against live power, thermal, and SLA signals. - Design memory and model-state lifecycle management for on-device LLM/SLM serving, including KV-cache and context-state caching, eviction, paging, and reuse across sessions. - Evaluate and architect around memory bandwidth and data-movement constraints, including high-bandwidth memory and on-chip interconnects, to maximize AI throughput and efficiency. - Independently research, define requirements, develop hands-on prototypes, and participate in design reviews to commercialize solutions. - Contribute to short- and long-term strategic direction for product and technology roadmaps by evaluating AI and Mobile industry trends.

What you'll need

- Bachelor's degree in Computer Science, Engineering, Information Systems, or related field with 8+ years of Hardware Engineering, Software Engineering, Systems Engineering, or related work experience (or Master's degree with 7+ years, or PhD with 6+ years). - Experience with Mobile SW architecture, deployment, system SW performance optimization, and benchmarking. - Experience with performance breakdown analysis across CPU, GPU, and NPU.

Nice to have

- Master's or PhD in Computer Science or Electrical Engineering, with 10 years of experience in High-Performance Computing System Engineering or Software and 5 years in AI system optimization. - Understanding of AI frameworks such as TensorFlow and PyTorch, as well as foundational models, small language models (SLMs), LoRA/adapter-based fine-tuning, and on-device personalization techniques. - Experience architecting SW frameworks for heterogeneous or multi-processor systems, graph-based (DAG) schedulers, and resource allocation across concurrent AI workloads. - Experience with LLM/SLM serving internals, including KV-cache management, paging/eviction strategies, and model-state reuse across sessions. - Understanding of memory subsystem and interconnect bandwidth (HBM/LPDDR, cache hierarchies, NPU memory architecture) and modeling AI networks for power and performance estimation. - Knowledge of the Android framework, Mobile platforms, and proficiency in programming languages including C, C++, and Python.

Skills: Snapdragon Mobile AI software, performance and power enhancements, SW and HW stacks, CPU, GPU, NPU, LLMs/SLMs, heterogeneous compute placement