Confluence Chatbot with GPT-3.5-turbo and Tokenizer

This Streamlit application allows users to interact with a chatbot that retrieves information from a selected Confluence space and generates responses using the OpenAI GPT-3.5-turbo model.

Features

• Confluence Integration: Connects to a Confluence instance to fetch page content and metadata.
• GPT-3.5-turbo Model: Leverages the power of OpenAI's language model for generating contextually relevant and informative responses.
• Tokenization: Efficiently manages text input to stay within the model's token limits.
• Chat History: Maintains a conversation history for reference.
• Space Selection: Allows users to choose from available Confluence spaces.
• Relevant Links: Provides links to relevant Confluence pages alongside chatbot responses.
• Error Handling and Logging: Includes error handling and logging for improved robustness.

Requirements

• Python 3.x
• Streamlit
• tiktoken
• atlassian-python-api
• openai
• numpy
• pandas
• beautifulsoup4
• nltk
• python-dotenv

Setup

1. Install dependencies:

   pip install streamlit tiktoken atlassian-python-api openai numpy pandas beautifulsoup4 nltk python-dotenv

2. Set up environment variables:

   • Create a .env file in your project directory.

   • Add the following variables with your actual values:

     openai-api-key = <your-openai-api-key>
     confluence-url = <your-confluence-url> (e.g., https://your-domain.atlassian.net/wiki)
     confluence-username = <your-confluence-email>
     confluence-api-token = <your-confluence-api-token>
     

3. Run the application:

   streamlit run your_script_name.py

Usage

1. Select a Confluence space from the dropdown.
2. Enter your query in the text area.
3. Click "Get Chatbot Response" to receive an answer based on the Confluence content.
4. View the response and relevant links below.
5. Continue the conversation by asking more questions.

Notes

• Ensure you have the necessary permissions to access the Confluence space.
• The chatbot's knowledge is limited to the content within the selected space.
• The application handles token limits to avoid exceeding the model's capacity.
• Refer to the logging output for detailed information on the application's behavior.
• Consider customizing the prompt construction and response filtering for more tailored results.