A Retrieval-Augmented Generation (RAG) Approach to Toothpaste Ingredient Analysis
Afanasev, Petr (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025121235397
https://urn.fi/URN:NBN:fi:amk-2025121235397
Tiivistelmä
This thesis focuses on researching and developing a prototype RAG (Retrieval-Augmented Generation) application for toothpaste and ingredient analysis, which gives users deeper insights about each toothpaste, helps them understand its potential impacts on oral and overall health, and simplifies everyday decision-making.
Initially, the framework and application architecture were designed and chosen. It was decided to use a Python server, a Node.js server, a Next.js frontend, and a PostgreSQL database. The development process included writing the frontend JavaScript code, creating a parser for extracting external data, implementing a Node.js server for core tasks, implementing a Python server for AI tasks, such as handling embeddings and image processing, and configuring the database to store the knowledge about ingredients and toothpaste products.
The result was a new web application tool that allows users to analyse toothpastes, add new products to the database, and get information about toothpastes. Basic functionalities, such as adding new toothpaste records, viewing existing toothpastes, and performing search and filtering operations, were fully implemented. However, several potential future enhancements were identified, including updating the server infrastructure, implementing a proxy for the deployed server, and other improvements.
Initially, the framework and application architecture were designed and chosen. It was decided to use a Python server, a Node.js server, a Next.js frontend, and a PostgreSQL database. The development process included writing the frontend JavaScript code, creating a parser for extracting external data, implementing a Node.js server for core tasks, implementing a Python server for AI tasks, such as handling embeddings and image processing, and configuring the database to store the knowledge about ingredients and toothpaste products.
The result was a new web application tool that allows users to analyse toothpastes, add new products to the database, and get information about toothpastes. Basic functionalities, such as adding new toothpaste records, viewing existing toothpastes, and performing search and filtering operations, were fully implemented. However, several potential future enhancements were identified, including updating the server infrastructure, implementing a proxy for the deployed server, and other improvements.