Using the Knowledge Base
LangBot natively supports RAG (Retrieval-Augmented Generation). You can create a built-in knowledge base and bind it to a pipeline, allowing the pipeline to answer questions based on the contents of the knowledge base. LangBot also supports connecting to external knowledge retrieval services (e.g., Dify, RAGFlow).
Build a Built-in Knowledge Base
INFO
- Creating and using a built-in knowledge base requires configuring an embedding model first. Please read Configure Models.
On the knowledge base page, click the Create Knowledge Base button, fill in the name, select an already configured embedding model, then click Create.
After creation, go to the "Documents" tab of your knowledge base and upload documents. LangBot will automatically parse and index these in the background.
Connect External Knowledge Base
On the knowledge base page, switch to the "External Knowledge Base" tab and add an external retriever. In the dialog, select the knowledge retriever plugin to use (please install the related plugin in Plugin Management first). Supported retrievers can be found in the Plugin Marketplace.
Use the Knowledge Base
Go to your pipeline configuration, under the "AI" tab, choose Local Agent as the runner, then select the knowledge base you just created.
INFO
LangBot's built-in knowledge base can be used only when the runner is set to Local Agent. For other runners, refer to their documentation.
Now you can use the knowledge base in Chat Debug, or through the bot linked to this pipeline:
