Intern – AI Model Efficiency (Quantization) System/Research Engineer
Core
Developing AI technology for edge devices to run models (LLMs, VLMs) efficiently via quantization and model compression.
Role type
Research intern (AI Model Efficiency)
Builds
Production-ready solutions for next-generation edge devices
Domain
AI/ML, Edge Computing, Hardware Optimization
Deliverable
production ML models
Required skills
Python, Deep Learning frameworks (PyTorch), Quantization methods (PTQ/QAT), Model profiling, Debugging
Preferred skills
Triton kernels, CUDA programming, Model compression techniques, Software design fundamentals
Technologies
PyTorch, Triton, CUDA
Responsibilities
Implement and benchmark low-bit quantization methods; Contribute to evaluation and profiling pipelines; Prototype and optimize efficient inference kernels; Collaborate with global teams to move prototypes to production; Document findings and present results.
