Accommodation Finder using RAG
A Gradio-based chatbot application designed to help users find apartments. It leverages a Retrieval-Augmented Generation (RAG) pipeline implemented using LangChain to provide apartment suggestions based on mimicked scraped data from real estate website due to terms of service restrictions.
Key Features:
User Input:
Users interact with a chat-based interface to ask about available apartments , specifying any preferences they might have (e.g., bedrooms, budget).
Dynamic Response Generation:
The application dynamically retrieves information from a Chroma-based vector store, powered by OpenAI embeddings, to provide relevant apartment listings.
The results are presented in two stages:
A brief summary that directly addresses the user's preferences.
A well-formatted table containing details such as apartment name, number of bedrooms, rent, location, and availability.
Real-time Interaction:
The chatbot delivers responses in real-time, making it easy for users to quickly find and evaluate available accommodation options.
Custom Prompting:
A custom prompt is employed to ensure that the output is clear, relevant to the user’s needs, and neatly formatted with both a description and tabular data.
Scraped Data Integration:
Apartment listings are retrieved from scraped data format, enhancing the application's ability to provide up-to-date and accurate suggestions.
Skills:
#RAG #LangChain #Gradio #OpenAI_API #WebScraping #Python