n8n API Schema Extractor Workflow

This n8n workflow is designed to extract and process API schemas, likely from a Google Sheet, and then use AI models (Google Gemini) for various tasks such as classification and information extraction. It leverages Langchain nodes for advanced text processing and Qdrant as a vector store.

What it does

This workflow performs the following key steps:

1. Manual Trigger: The workflow is initiated manually.
2. Google Sheets Integration: It interacts with Google Sheets, likely to read or write API schema data.
3. HTTP Request: Makes an HTTP request, probably to fetch API schema definitions from a URL.
4. Data Transformation (Edit Fields): Modifies or sets fields within the incoming data.
5. Loop Over Items: Processes items in batches, indicating that it handles multiple schema entries.
6. Conditional Logic (If): Introduces branching logic based on certain conditions.
7. Switch Node: Further branches the workflow based on different cases or values.
8. Wait: Pauses the workflow for a specified duration.
9. Code Execution: Executes custom JavaScript code for specific data manipulation or logic.
10. Filter: Filters items based on defined criteria.
11. Execution Data: Accesses or manipulates execution-related data.
12. Recursive Character Text Splitter: Splits text into smaller, manageable chunks, a common step in preparing data for AI models.
13. Default Data Loader: Loads documents into a standardized format for Langchain processing.
14. Aggregate: Combines multiple items into a single output.
15. Remove Duplicates: Eliminates duplicate entries from the data.
16. Split Out: Splits out nested data structures into separate items.
17. Qdrant Vector Store: Interacts with a Qdrant vector database, likely for storing and retrieving vector embeddings of the API schemas.
18. Embeddings Google Gemini: Generates vector embeddings using the Google Gemini model.
19. Google Gemini Chat Model: Utilizes the Google Gemini chat model for conversational AI tasks.
20. Text Classifier: Classifies text based on predefined categories using AI.
21. Information Extractor: Extracts structured information from unstructured text using AI.
22. Google Drive Integration: Interacts with Google Drive, potentially for storing processed schemas or reports.
23. Execute Sub-workflow: Calls another n8n workflow as a sub-process.
24. Execute Workflow Trigger: Allows this workflow to be triggered by another workflow.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance.
• Google Sheets Account: Configured credentials for Google Sheets.
• Google Drive Account: Configured credentials for Google Drive.
• Qdrant Instance: Access to a Qdrant vector database.
• Google Gemini API Key/Access: Credentials for Google Gemini (for Embeddings and Chat Model).
• Langchain Nodes: Ensure the @n8n/n8n-nodes-langchain package is installed and enabled in your n8n instance.

Setup/Usage

1. Import the workflow: Download the JSON content and import it into your n8n instance.
2. Configure Credentials:
   • Set up your Google Sheets credentials.
   • Set up your Google Drive credentials.
   • Configure your Google Gemini API key/access for the Embeddings and Chat Model nodes.
   • Configure your Qdrant Vector Store connection details.
3. Customize Nodes:
   • Adjust the "Google Sheets" node to point to your specific spreadsheet and data.
   • Modify the "HTTP Request" node with the API endpoint you wish to process.
   • Review and update the logic in the "If" and "Switch" nodes to match your specific schema processing requirements.
   • Customize the "Edit Fields" and "Code" nodes for any specific data transformations.
   • Configure the "Text Classifier" and "Information Extractor" nodes with the desired labels and schema for AI processing.
4. Execute the workflow: Click the "Execute workflow" button on the "Manual Trigger" node to run the workflow.