LLM-Curated Catalog at Amazon

As an AI Engineer at Amazon, I led the development of a comprehensive LLM-curated catalog aimed at assisting customers in achieving their objectives through a diverse array of ideas and missions.

Key Achievements

Human-in-the-Loop Framework

Incorporated a Human-in-the-Loop (HITL) framework to continuously fine-tune and improve the LLM-curated catalog. This involved leveraging expert feedback to refine prompts and model outputs, ensuring the highest quality and relevance of the generated content.

Automated Data Pipeline

Developed an automated data pipeline and processes to streamline expert reviews and feedback loops, integrating this feedback into the model training workflow to enhance the accuracy and utility of the catalog.

Customer Experience Enhancement

Led the initiative to showcase the LLM-curated catalog on Amazon.com, highlighting its potential to provide personalized recommendations and insights, thereby enhancing customer experience and driving engagement.

Technologies Used

• Large Language Models
• Human-in-the-Loop AI systems
• Automated data pipelines
• Machine learning workflows