Note: The job is a remote job and is open to candidates in USA. NVIDIA is a leading company in AI infrastructure, and they are seeking a Senior AI Compiler Engineer to join their team. The role focuses on developing innovative AI compiler solutions and optimizing low-level GPU programming through AI-based technologies.

Responsibilities

• Help trailblaze company efforts in applying AI within conventional compilation pipelines
• Design and implement AI-based technology addressing core problems of low-level GPU programming
• Build training pipelines for supervised fine-tuning and reinforcement learning (RL/RLHF-style or policy optimization variants)
• Define model inputs/outputs over compiler low level compiler representations
• Develop evaluation frameworks to measure code quality, runtime, compile-time overhead, and correctness
• Intelligent (domain` task based) prompt engineering
• Collaborate with compiler engineers to integrate learned policies into production toolchains
• Prototype and iterate on model architectures, prompts, and fine-tuning strategies for scheduling and allocation tasks
• Create datasets from compiler traces, optimization passes, and target-specific performance signals
• Apply RL techniques to optimize for downstream objectives (performance, spill reduction, instruction-level parallelism, etc.) and run rigorous experiments, ablations, and benchmarking across workloads and hardware targets

Skills

• M.S./PhD degree in Computer Engineering, Computer Science related technical field (or equivalent experience)
• 5+ years of experience building AI/ML systems
• Strong software engineering skills in Python and at least one systems language (C++ preferred)
• Hands-on experience training/fine-tuning large models (Transformers, PEFT/LoRA, distributed training)
• Solid understanding of machine learning fundamentals and experimentation best practices
• Experience with reinforcement learning (e.g., policy gradients, actor-critic, offline RL, bandit-style optimization)
• Knowledge of prompt-engineering techniques
• Ability to work across research and engineering, from prototype to production
• Distributed training/inference at scale
• Experience working with the NVIDIA NeMo framework
• Understanding of GPU performance, experience with benchmarking suites and performance profiling tools
• Formal methods or static analysis familiarity for correctness guarantees
• CUDA programming experience

Benefits

• Equity
• Benefits

Company Overview

Company H1B Sponsorship