Getting Started#

RAGFlow on DKubeX turns your documents into knowledge bases you can chat with. This guide takes you from signing in to your first answer.

Sign in#

RAGFlow uses DKubeX single sign-on. Open the app from your DKubeX dashboard and you are signed in automatically with your existing credentials — no separate RAGFlow login.

Set up the DKubeX model provider#

RAGFlow needs models to embed your documents and generate answers. On DKubeX these come from the built-in DKubeX provider, backed by SecureLLM.

  1. Click your avatar in the top-right and open the Model providers page.

  2. Find the DKubeX provider, start its setup, paste your SecureLLM API key, and confirm. RAGFlow registers the chat, embedding, and rerank models your key permits.

  3. Open Set Default Models and choose your default chat and embedding models.

See Using models on DKubeX for details.

Create your first knowledge base#

  1. Create a dataset.

  2. On its Configuration page, choose a chunking method (for example, General) and your embedding model.

  3. Upload one or more documents and parse them.

  4. Optionally run a retrieval test to confirm the right chunks come back.

See Knowledge bases for the full set of parsing and tuning options.

Start a chat#

  1. Click the Chat tab, then Create an assistant.

  2. Give the assistant a name and select the dataset you just created.

  3. Keep Show quote enabled so answers cite their sources, and start chatting.

See Chat for all assistant settings.

Next steps#