Location
London

Hours
Full Time

Salary
Competitive, commensurate with experience

About the Role
We are seeking a passionate, talented, and inventive Applied Scientist with a strong machine learning background to join the Alexa for Shopping Science team. This role focuses on building industry-leading language technology powering Alexa for Shopping, an AI-driven search and shopping assistant that supports customers throughout their shopping journey. You will develop and optimize language model-powered conversational experiences, leveraging large and small language models (LLM/SLM) through instruction design, contextual grounding, model fine-tuning, evaluation frameworks, and experimentation to improve quality, robustness, and customer impact.

Your work will combine scientific rigor with product intuition to raise the bar for conversational AI performance at Amazon scale. The mission is to simplify product discovery and shopping by assisting with product research, comparisons, recommendations, answering questions, enabling shopping from images or videos, and providing visual inspiration. You will apply advanced analytics, Natural Language Processing (NLP), Machine Learning (ML), A/B testing, causal inference, and data-driven insights to continuously enhance our systems.

Key responsibilities include developing and maintaining LLM agents, automated evaluation pipelines, rubric design, and dataset curation to measure nuanced response quality. You will collaborate across teams to experiment with retrieval augmentation, context enrichment, prompt decomposition, and model fine-tuning or post-training strategies. Where latency and cost constraints apply, you will lead post-training of small language models including supervised fine-tuning, preference optimization, and distillation to deliver low-latency conversational shopping experiences.

You will design and evaluate agentic architectures balancing diverse shopping use cases, making principled choices across single-agent and multi-agent systems, memory management, and tool orchestration to optimize quality, latency, and reliability at scale. Leveraging petabytes of data, you will identify opportunities to improve conversational system performance.

A typical day involves hands-on analysis of large-scale multimodal interaction datasets, using statistical methods and experimentation to develop scalable evaluation and optimization approaches for LLM-based shopping assistants. You will conduct deep-dive analyses to improve conversational relevance, grounding, customer satisfaction, and business impact. Collaboration with product management and engineering teams will translate analytical insights into production systems. You will communicate results effectively to both technical and non-technical audiences through presentations, reports, and data visualizations.

The Alexa for Shopping Science team, based in London, includes approximately 150 engineers, designers, and product managers focused on advancing AI-driven shopping experiences. The team works on conversational AI systems that enable agentic behavior, multimodal query understanding, and personalized shopping guidance using state-of-the-art NLP, generative AI, information retrieval, machine/deep learning, and data mining techniques. We actively participate in scientific communities to validate and advance our work.

Experience
- PhD or Master's degree with experience in Computer Science, Computer Engineering, Machine Learning, or related fields
- Expertise in state-of-the-art deep learning model architecture design, training, optimization, and model pruning
- Proven track record of patents or publications in top-tier peer-reviewed conferences or journals
- Proficiency in programming languages such as Java, C++, Python, or related languages
- Experience in algorithms and data structures, parsing, numerical optimization, data mining, parallel and distributed computing, or high-performance computing
- Experience building machine learning models for business applications

About You
- Hands-on expert in large language model post-training, including supervised fine-tuning and large-scale reinforcement learning
- Strong understanding of algorithms for large-scale distributed training
- Demonstrated strong publication record in top-tier NLP/LLM conferences such as NeurIPS, ICLR, ICML, EMNLP, ACL, NAACL with 500+ citations
- Passionate about advancing conversational AI and delivering impactful shopping experiences at scale
- Collaborative team player with excellent communication skills for technical and non-technical audiences

Qualifications
- Advanced degree (PhD preferred) in relevant technical fields
- Deep knowledge of machine learning, NLP, and conversational AI technologies
- Ability to design and implement scalable evaluation frameworks and experimentation methodologies
- Experience working with large-scale datasets and distributed computing environments