Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.langbot.app/llms.txt

Use this file to discover all available pages before exploring further.

LangBot natively supports RAG (Retrieval-Augmented Generation). You can create a knowledge base and bind it to a pipeline, allowing the pipeline to answer questions based on the contents of the knowledge base. Knowledge bases are powered by Knowledge Engine plugins, with different Knowledge Engines providing different indexing and retrieval strategies. You can find available Knowledge Engine plugins in the Plugin Marketplace.

​
Creating a Knowledge Base

On the knowledge base page, click the Create Knowledge Base button:
  1. Fill in the knowledge base name
  2. Select a Knowledge Engine (provided by installed plugins)
  3. Fill in the relevant parameters based on the selected engine’s configuration form (e.g., embedding model, chunk size, etc.)
  4. Click the Create button
Different Knowledge Engines have different configuration parameters, depending on the engine’s creation_schema definition. Some engines may require configuring an embedding model first. Please read Configure Models.
If the selected Knowledge Engine supports document upload (declares DOC_INGESTION capability), after creation go to the “Documents” tab of your knowledge base and upload documents. LangBot will automatically parse and index these in the background.

​
Using 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. use_kb use_kb

​
Configure Reranking (Optional)

After selecting a knowledge base, you can configure a Rerank Model below to improve retrieval quality:
  • Rerank Model: Select a configured rerank model (if none available, add one in Model Configuration first)
  • Rerank Top-K: Number of most relevant results to keep after reranking, default is 5
Rerank models are optional. When configured, candidate documents from knowledge base retrieval are re-scored by the rerank model, keeping only the top-K most relevant results before passing them to the LLM, significantly reducing noise from irrelevant content.
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: use_kb_in_chat

​
Developing Custom Knowledge Engines

If existing Knowledge engine plugins don’t meet your needs, you can develop your own Knowledge engine plugin. See Component: Knowledge Engine for development details.