Overview

This node operation creates a new RAG (Retrieval-Augmented Generation) resource by uploading a file and specifying metadata such as name, description, and chunking options. It is useful for users who want to build a RAG base from documents for enhanced information retrieval and AI-assisted querying. For example, uploading a PDF document with specific chunking settings to create a searchable knowledge base.

Use Case Examples

1. Uploading a PDF file named 'Research Paper' with a description and semantic chunking options to create a RAG base for AI querying.
2. Creating a RAG base from a text file with fixed-size chunking to enable efficient document search and retrieval.

Properties

Output

JSON

• id - Identifier of the created RAG resource.
• message - Response message from the API after creating the RAG.
• name - Name of the created RAG resource.
• description - Description of the created RAG resource.
• chunking_method - Chunking method used for the RAG base.
• chunk_size - Chunk size used for the RAG base.
• chunk_overlap - Chunk overlap used for the RAG base.
• separator - Separator used for fixed size chunking method.
• separators - Separators used for recursive chunking method.

Dependencies

• Straico API with authentication token

Troubleshooting

Error 'No binary data property "" found on item!' indicates the specified file binary data is missing. Ensure the file is correctly passed as binary input to the node.

• Invalid chunking options or missing required chunking parameters may cause API errors. Verify chunking options are correctly set according to the chunking method.
• API request failures may occur due to network issues or invalid API credentials. Check the Straico API key and network connectivity.
• If the API response format changes, the node may fail to parse the response. Verify the API response structure if unexpected errors occur.