Applied Scientist II, Alexa for Shopping Science UK
Core
Developing and optimizing language model-powered (LLM/SLM) conversational experiences for an AI-driven shopping assistant, focusing on instruction design, context engineering, and model fine-tuning to improve quality and robustness.
Role type
Senior Applied Scientist (LLM/SLM)
Builds
LLM agents, automated evaluation pipelines, and low-latency conversational shopping systems
Domain
E-commerce, Conversational AI, Natural Language Processing, Machine Learning
Deliverable
production ML models
Required skills
Deep learning model architecture design, model training and optimization, model pruning, algorithm design, data mining, parallel and distributed computing, high-performance computing, Python programming, LLM post-training (SFT, RL), prompt engineering, agentic system design, multimodal understanding
Preferred skills
Expertise in large-scale distributed training, reinforcement learning, publications in top-tier NLP/LLM conferences (NeurIPS, ICLR, ICML, EMNLP, ACL, NAACL) with 500+ citations
Technologies
Java, C++, Python, LLMs, SLMs, Retrieval Augmented Generation, A/B testing frameworks
Responsibilities
Develop and maintain LLM agents including automated eval pipelines and rubric design; lead post-training of small language models for low-latency experiences; apply deep learning techniques for ranking, relevance, and personalization; design and evaluate agentic architectures; analyze large-scale multimodal interaction datasets to improve response quality.
Seniority
Senior, hands-on IC
