← Back to Case Studies

Vietnamese Recipe Chatbot with RAG

Seedcom R&D | Nov 2020 - Oct 2024

Google GeminiRAGPythonFastAPI

Key Results

  • RAG system with customer basket history
  • Personalized culinary suggestions
  • Enhanced user engagement metrics
  • FastAPI deployment for high-throughput

The Problem

An e-commerce grocery platform wanted to increase customer engagement by offering personalised recipe suggestions based on items in their shopping basket. Generic recipe databases failed to account for customer preferences, dietary restrictions, and available ingredients.

The Approach

Built an intelligent chatbot using Retrieval-Augmented Generation powered by Google Gemini. The system indexes a curated Vietnamese recipe database and retrieves relevant recipes based on the customer’s current basket contents and purchase history. The RAG pipeline ensures responses are grounded in real recipes while allowing natural conversational interaction. Deployed via FastAPI for high-throughput serving.

The Results

The chatbot delivered personalised culinary suggestions that matched customer basket contents, significantly enhancing user engagement metrics. The RAG architecture ensured accurate, contextual recommendations while the FastAPI deployment handled production traffic efficiently.