AI Research Scientist - On-Device Large Language Models

Qualcomm seeks an AI Researcher focused on on-device LLM efficiency to advance next generation wireless and mobile platforms. You will research and prototype algorithms for model compression, quantization, pruning, distillation, and on device inference optimization across smartphones, IoT, and automotive systems. Collaborating with cross functional silicon, software, and product teams, you will design experiments, benchmark models, and translate research into production ready solutions. This role offers broad impact on 5G and edge AI, strong publication opportunities, and growth in a fast paced, innovation driven environment.

Responsibilities

• Research and develop methods for on-device LLM efficiency, including compression, quantization, pruning, and distillation
• Prototype and benchmark LLM inference on Qualcomm mobile, Io
• T, and automotive platforms
• Collaborate with silicon, software, and product teams to translate research into deployable solutions
• Design and run experiments to evaluate accuracy, latency, power, and memory trade-offs
• Publish and present findings internally and externally to help shape Qualcomm's AI roadmap
• Contribute to tooling and pipelines for efficient model deployment on edge devices

Required Skills

• Large language models (LLMs)
• Model compression and quantization
• Network pruning and distillation
• On-device and edge AI optimization
• Deep learning frameworks (Py
• Torch, Tensor
• Flow)
• C++ and Python programming
• GPU/NPUs and hardware-aware MLPerformance profiling and benchmarking
• Distributed training and experimentation
• MLOps and deployment pipelines