Machine Learning Research Engineer

Quick Facts

• Role: Machine Learning Research Engineer (LLM optimization for Azure production environments)

Description

Design and develop novel quantization and numerics algorithms and supporting tooling to enable efficient LLM inference and training in Microsoft Azure. Analyze performance bottlenecks in state-of-the-art quantized LLMs, prototype emerging low-precision data formats on accelerator hardware, and co-design model architectures optimized for low-precision deployment with AI/ML teams. Collaborate cross-functionally to align on accuracy and performance goals and ensure end-to-end system efficiency.

Responsibilities

• Design and develop quantization and numerics algorithms for efficient LLM deployment

• Build proof-of-concept software and model optimization tooling to streamline quantized model deployment

• Analyze and improve performance bottlenecks in quantized LLM architectures

• Prototype and evaluate low-precision data formats using accelerator SDK proof-of-concepts

• Co-design model architecture optimized for low-precision deployment with AI/ML teams

• Collaborate with data scientists and ML researchers/engineers on accuracy and performance targets

• Partner with hardware architecture and AI software framework teams for end-to-end system efficiency

Requirements

• Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering (or related) plus the required years of related experience as specified in the posting

• Ability to pass Microsoft Cloud background check upon hire/transfer and every two years thereafter (security screening requirement)

Benefits

• Eligible for benefits and other compensation (subject to role/location)

• Typical U.S. base pay range: USD $102,100–$202,200 per year (different ranges for specific metro areas)