Hi there !
Here’s the situation :
I’ve set up a workflow in n8n to analyze a binary file (a CV in PDF format) using several OpenAI nodes. Here’s how it works right now:
- Retrieving the CV: An Airtable node fetches the CV file in binary format.
- Converting the PDF to an Image: An HTTP Request node converts the PDF to an image (PNG).
- Analyzing the Binary File: The converted file is then passed through several OpenAI nodes to extract various information.
The issue I’m facing is that while the first OpenAI node works correctly, the subsequent nodes can’t access the binary file. I want to avoid making another API call to retrieve the document before each OpenAI node.
What do I want to achieve :
- Binary Data Propagation: I want the binary data to be available for each OpenAI node without having to make another API call to retrieve the document.
- Condition on the First OpenAI Node: If the first OpenAI node returns an error message indicating that the document is not a valid or analyzable CV, I want the other OpenAI nodes to not be executed.
Here’s a simplified example of the workflow I try to create :
If anyone has any ideas on how to make this work, I’d love to hear them. I’ve already tried several potential solutions, but nothing has worked so far.
Thank you folks ! ![]()
- n8n version: 1.51.2
- Database: default (SQLite)
- n8n EXECUTIONS_PROCESS setting (default: own, main): (I Don’t
- Running n8n via (Docker, npm, n8n cloud, desktop app): Docker
- Operating system: Ubuntu