Quickstart
Get to know RAG-DL by building an OpenAI-powered chatbot application.
Prerequisites
Open RAG-DL and start a new workflow
From the RAG-DL dashboard, click create a new workflow. A blank workspace opens where you can build your flow.
Tip: If you don't want to create a blank workflow, you can import an existing one by clicking import a new workflow.
Build the Basic Prompting Workflow
Follow these steps to create a simple chat workflow with OpenAI integration:
Add Components
- From the sidebar, locate Chat → Chat Input and drag it to your canvas. This component handles user messages.
- Add a Prompts → Prompt component. This will store your system instructions for the AI.
- Add Chat → Chat Output. This displays responses in the chat interface.
- Finally, add Models → OpenAI. This component manages the API communication.
Connect Components
Components connect from left (inputs) to right (outputs). Connect them as follows:
- Chat Input's Message → OpenAI's Input
- Prompt's Prompt Message → OpenAI's System Message
- OpenAI's Text → Chat Output's Text
Note: Hover over connection ports to see compatible data types. For detailed information about component connections, see our Components Overview guide.
Final Setup
Before running your workflow:
- Configure your OpenAI API key in the OpenAI component
- Add system instructions in the Prompt component
Your completed Basic Prompting workflow should look like this:
Once configured, you can start chatting with your AI assistant:
