Enhancing the shopping experience on Amazon through the conversational capabilities of large language models
Requirements
- Experience with Machine and Deep Learning toolkits such as MXNet, TensorFlow, Caffe and PyTorch
- Experience with CUDA, cuDNN, cuBLAS and other GPU kernel-level optimization techniques
- 5+ years of programming with at least one software programming language experience
- Experience with Neuron hardware (Inferentia and Trainium chips) and NKI kernel optimization
- 3+ years of hands-on experience with CUDA programming and GPU kernel development
- Experience leading design or architecture (design patterns, reliability and scaling) of new and existing systems
- Experience with full software development life cycle, including coding standards, code reviews, source control management, build processes, testing, and operations
Responsibilities
- Lead projects in architecting, designing, developing, and optimizing high-performance kernel implementations for large language model
- Guide your team in creating and optimizing innovative kernels, custom operators, and low-level optimizations that maximize hardware utilization and minimize computational overhead
- Establish best practices for kernel development, memory management, and parallel computing that dramatically reduce inference latency and boost throughput for transformer-based models
- Develop kernel fusion techniques, attention mechanism optimizations, and matrix multiplication accelerations at scale
- Partner with engineers and scientists in a fast-paced environment to deliver measurable performance gains
- Drive technical roadmap, performance benchmarking, and optimizations focused on kernel-level improvements
- Optimize the computational heart of AI systems
Other
- 5+ years of non-internship professional software development experience
- Experience as a mentor, tech lead or leading an engineering team
- Bachelor's degree in computer science or equivalent
- Ability to work in a fast-paced environment
- Ability to deliver measurable performance gains
