Skip to content

Latest commit

 

History

History
121 lines (81 loc) · 3.89 KB

README.md

File metadata and controls

121 lines (81 loc) · 3.89 KB

Google Reviews AI Query System

This open-source project combines cutting-edge technologies to enable the retrieval, summarization, and querying of Google Maps customer reviews. By leveraging the Outscraper API and OpenAI's GPT models, the system integrates a Retrieval-Augmented Generation (RAG) approach, utilizing the Llama_Index library for efficient document indexing and query handling. Through embeddings, it ensures precise semantic search capabilities, offering users a robust tool to analyze reviews and extract actionable insights. The system fetches reviews from Google Maps, processes them into a structured format, and enables interactive querying using advanced AI models. This allows users to analyze reviews, extract key insights, and answer specific questions seamlessly.

Features

  • Outscraper Integration: Automatically fetch customer reviews for the last month from Google Maps using Outscraper.
  • OpenAI-Powered Query Engine: Summarize and query review data using OpenAI's GPT models.
  • Interactive Query System: Engage in a conversational interface to retrieve insights from review data.

Installation

Prerequisites

  1. Python 3.8 or higher.
  2. An OpenAI API key.
  3. An Outscraper API key.
  4. Install required Python libraries: 
    pip install -r requirements.txt

Setup

  1. Clone the repository:

    git clone <repository_url>
cd <repository_name>

  2. Create a .env file to store sensitive information:

    OPENAI_API_KEY=<your_openai_api_key>
OUTSCRAPER_API_KEY=<your_outscraper_api_key>
CUSTOMER_PLACE_ID=<place_id>

  3. The .env file and avaliacoes.json (reviews data) are ignored by Git to protect sensitive customer data.

  4. Install any additional dependencies:

    pip install -r requirements.txt

Project Structure

.
├── Agent.py          # Query engine for document summarization and context retrieval
├── avaliacoes.json   # Review data fetched from Outscraper (ignored in .gitignore)
├── main.py           # Main script to interact with the system
├── outscraper.py     # Functions for fetching reviews using the Outscraper API
├── prompts.json      # Contains system prompts for the query engine
├── secret.py         # Handles secure retrieval of API keys
├── unix.py           # Utility to calculate timestamps
├── .env              # Stores API keys (ignored in .gitignore)
├── .gitignore        # Specifies ignored files

Usage

  1. Run the System:

    python main.py

  2. The program works as follows:

    • Fetches Google Maps reviews for the last month using Outscraper.
    • Saves the reviews in a JSON file (avaliacoes.json).
    • Processes the reviews to allow interactive querying and summarization using OpenAI's GPT models.
    • Allow you to ask questions interactively, such as:
      • "Summarize the customer feedback."
      • "What are the most common complaints?"

  3. Example:

    Voce: Summarize the reviews for this month.
Seu agente de IA: The reviews highlight excellent customer service and fast delivery but mention occasional issues with product quality.

Configuration

  • Prompts: Customize the prompts.json file to adjust the behavior of the query engine.
  • Ignored Files: The .env and avaliacoes.json are excluded from version control to protect sensitive data.

Notes

  • This system relies on the availability of Outscraper and OpenAI APIs.
  • Ensure that API usage limits are sufficient for your needs.

Contributions

Contributions are welcome! Please create an issue or submit a pull request to suggest improvements or report bugs.

Disclaimer

This project is designed for educational and informational purposes. Use responsibly and ensure compliance with local regulations when processing user data.