Applied Scientist II, SSG Science
Core
Develop and optimize Gen AI models for edge platforms on Amazon devices (Kindle, Fire, Echo, Astro) using custom ML hardware.
Role type
Applied Scientist (Edge AI & Model Optimization)
Builds
Next-generation edge models and production-ready ML solutions for consumer devices
Domain
Consumer Electronics / Edge AI / High-Performance Computing
Deliverable
production ML models
Required skills
Gen AI model optimization (quantization, pruning, distillation, fine-tuning), Neural Edge Engine architecture, deep learning theory, information theory, scientific computing, non-equilibrium thermodynamics, Java/C++/Python programming, algorithms and data structures, numerical optimization, parallel and distributed computing, high-performance computing
Preferred skills
Professional software development, full lifecycle management of ML models from development to deployment
Technologies
Neural Edge Engine, NeurIPS, ICLR, MLSys
Responsibilities
Quantize, prune, distill, and finetune Gen AI models for edge platforms; Analyze deep learning workloads to map them to Amazon's Neural Edge Engine; Train custom Gen AI models that beat SOTA; Collaborate with compiler engineers, hardware architects, and product teams; Publish in open source and present at key ML conferences
Seniority
Mid-Senior, hands-on IC
