Practical overview
When it fits
Use it when you want a documented starting point for a Confluence-based help-desk chatbot and can configure the required Python environment, Confluence credentials and OpenAI key.
Who it is for
- Teams seeking a question-answering interface over company Confluence pages.
- Developers exploring a documented retrieval-augmented generation pipeline with Streamlit, LangChain and OpenAI components.
- Practitioners who want to evaluate RAG responses against topic-specific question-and-answer samples.
Documented uses
- Build a help-desk chatbot that answers questions from a Confluence space.
- Collect and split Confluence content into chunks for retrieval.
- Run an evaluation workflow using a customized question-and-answer dataset.
Requirements & limits
The repository is public and licensed under MIT. The documented setup requires Confluence credentials and an OpenAI API key.
- The documented content ingestion is for Confluence; another source requires modifying the loader.
- The setup calls for Python 3.10, repository dependencies, Confluence account details and API credentials, and an OpenAI API key.
- The README describes Markdown-aware and recursive character splitting, with Markdown preservation noted as useful for header-based splitting.
- The evaluation workflow expects users to replace the example dataset with their own topic questions and answers.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up and launch the Confluence chatbot
Prepare the documented Python environment and configure the Streamlit chatbot to use a Confluence space and OpenAI.
Before you start
- Python 3.10
- The repository and its requirements.txt dependencies
- A Confluence space containing page information
- A Confluence API key and the associated email address
- An OpenAI API key
Steps
Create and activate a Python 3.10 virtual environment, then install the dependencies from requirements.txt.
Copy .env.template to .env and check config.py and the environment template.
Enter the Confluence space key and name, the account email, and the Confluence and OpenAI API keys in the environment configuration.
From the src directory, run streamlit run streamlit.py.
Expected outcome
The documented Streamlit interface for asking questions using company Confluence data is launched.
Guide limitations
- The README describes Confluence as the content source; using another source requires modifying the ingestion step.
- The setup instructions do not establish that the application has been independently tested.
Evaluate RAG answers with a topic-specific dataset
Use the included evaluation script with questions and answers tailored to a topic.
Before you start
- The repository and its Python dependencies
- A topic-specific question-and-answer dataset to replace the example file in the data folder
Steps
Replace the evaluation dataset file in the data folder with topic-specific questions and answers.
From the src directory, run python evaluate.py.
Expected outcome
The included script evaluates the RAG model using the replacement question-and-answer samples.
Guide limitations
- The README describes evaluation against question-and-answer samples but does not specify evaluation metrics.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 15 October 2026
- Original website title
- BastinFlorian/RAG-Chatbot-with-Confluence · Official homepage title · retained in its source language