Qualcomm
Machine Learning Compiler Engineer
$224K–$335K
New York City, New York · On-site · Full-time · Posted Sep 17, 2026
- Experience
- Minimum qualifications: Bachelor's degree and 8+ years, Master's degree and 7+ years, or PhD and 6+ years of relevant work experience. The role spans entry to experienced levels.
- Education
- Bachelor's, Master's, or PhD in Computer Science, Engineering, Information Systems, or a related field; the posting separately asks for a degree in computer science or applied mathematics.
About the job
Develop and improve machine-learning compilers and optimization algorithms for Qualcomm accelerators, incorporating generative AI into compiler workflows and compiler capabilities.
What you'll do
- Improve compiler optimizations through benchmark analysis and profiling, and develop new compiler and optimization algorithms.
- Incorporate generative AI into compiler development workflows and compilers; upstream algorithms to open-source projects.
- Author research publications and represent Qualcomm at conferences and industry forums; senior engineers also lead projects while contributing technically.
What they ask for
- Experience with compiler development and computer architecture, plus ML experience.
- Experience with software engineering.
- A degree in computer science or applied mathematics.
- Strong intellectual ability, motivation, achievement history, and oral and written communication skills.
Nice to have
- Experience with MLIR, MLIR dialects (LinAlg, Affine), PyTorch 2.0, TVM, Triton, and/or LLVM.
- Experience in areas such as generative-AI-assisted compiler development, DSLs, ML optimization, high-performance computing, polyhedral optimization, or GPU programming.
Pay
$223,600.00 - $335,400.00 (pay range; salary)
Tools
Qualcomm Hexagon Neural Signal Processor, Adreno GPUs, LLVM, MLIR, Triton, Torch-MLIR, PyTorch 2.0, TVM
AI focus
Use generative AI to accelerate compiler development and build AI capabilities into compilers, including AI-assisted code generation, analysis, refactoring, and review.
Model-written summary · the original posting is authoritative