Senior Engineer- Machine Learning

Job Description

Join the exciting Generative AI team at Qualcomm focused on integrating cutting edge GenAI models on Qualcomm chipsets. The team uses Qualcomm chips’ extensive heterogeneous computing capabilities to allow inference of GenAI models on-device without a need for connection to the cloud. Our inference engine is designed to help developers run neural network models trained in a variety of frameworks on Snapdragon platforms at blazing speeds while still sipping the smallest amount of power. Utilize this power efficient hardware and Software stack to run Large Language Models (LLMs) and Large Vision Models (LVM) at near GPU speeds!

Responsibilities:

In this role, you will spearhead the development and commercialization of the Qualcomm AI Runtime (QAIRT) SDK on Qualcomm SoCs. As an AI inferencing expert, you'll push the limits of performance from large models. Your mastery in deploying large C/C++ software stacks using best practices will be essential. You'll stay on the cutting edge of GenAI advancements, understanding LLMs/Transformers and the nuances of edge-based GenAI deployment. Most importantly, your passion for the role of edge in AI's evolution will be your driving force.

Requirements:

• Master’s/Bachelor’s degree in computer science or equivalent.
• 2-4 years of relevant work experience in software development.
• Strong understanding of Generative AI models – LLM, LVM, LMMs and building blocks (self-attention, cross attention, kv caching etc.)
• Floating-point, Fixed-point representations and Quantization concepts.
• Experience with optimizing algorithms for AI hardware accelerators (like CPU/GPU/NPU).
• Strong in C/C++ programming, Design Patterns and OS concepts.
• Good scripting skills in Python.
• Excellent analytical and debugging skills.
• Good communication skills (verbal, presentation, written).
• Ability to collaborate across a globally diverse team and multiple interests.

Preferred Qualifications

• Strong understanding of SIMD processor architecture and system design.
• Proficiency in object-oriented software development and familiarity
• Familiarity with Linux and Windows environment
• Strong background in kernel development for SIMD architectures.
• Familiarity with frameworks like llama.cpp, MLX, and MLC is a plus.
• Good knowledge of PyTorch, TFLite, and ONNX Runtime is preferred.
• Experience with parallel computing systems and languages like OpenCL and CUDA is a plus.