BookAssistant README

Overview

The BookAssistant is designed to interact with users, retrieve book information using various tools, and present structured responses. It leverages a language model (ChatAnthropic), Vector database, and a LangChain state graph to efficiently manage interactions and data flow.

Untitled.video.-.Made.with.Clipchamp.2.mp4

Components

1. Initialization and Setup

The BookAssistant class initializes several components:

• Language Model: Uses ChatAnthropic.
• Tools: Defines tools (googleAPI_retrieval, present_book_info, search_db) for retrieving and presenting book information.

2. Graph Construction

• State Graph: Constructs a state graph to manage state transitions and tool interactions.

  • Nodes and Edges:
    • Adds nodes for an assistant (handling user interactions) and tools (containing tools and error handling).
    • Establishes conditional edges and entry points (assistant) within the graph.

3. Tools

• googleAPI_retrieval: This class acts as a wrapper for retrieving data from the Google Books API. The fetched data is stored in both JSON format and a Chroma database, facilitating efficient extraction and search operations.

• present_book_info: This function extracts data from the JSON file and renders it in the chat interface using a token, enhancing data presentation reliability and reducing model token usage.

• search_db: Enables vector similarity search on retrieved data, enhancing search capabilities based on content similarity.