SakeSensei

• 

  SakeSensei-Architecture-Diagram

We are increasingly hearing from those in the sake industry (sake shops and breweries) about the difficulty of "how to select and recommend" sake. Although there are customers interested in sake, it is difficult to determine what to recommend to them. Many young people develop a dislike for sake after their first experience, often because they chose a bottle based on price or one that didn't suit their taste. This makes the initial recommendation particularly crucial. We hypothesized that we could solve this challenge by using an AI agent combined with RAG to provide recommendations based on "preference data and a knowledge base." In addition to helping with the initial sake selection, we decided to develop a web application incorporating an AI agent to continuously support the user's sake lifestyle.

Personalized Sake Recommendations: The AI suggests sake tailored to individual preferences. Tasting Records: Keep a record of the sake you've tried to create your own personal sake notebook.

Built With

• agentcore
• amazon-web-services
• bedrock
• strandsagent