Classify & extract data from floorplans with Mistral AI OCR & JigsawStack

by Stephan Koning ✓ · 317 views · created 2025-09-09 · Document Extraction, AI Summarization

What this workflow does

This workflow has 17 nodes and runs when an HTTP POST request hits its webhook URL. It uses only n8n's built-in nodes. It also calls api.jigsawstack.com over HTTP.

How it runs

Trigger: When an HTTP POST request hits its webhook URL.

1. Check – GDPR Consent (If): checks a condition and branches into true / false
2. Process – Multiple File Uploads (Code): runs custom JavaScript code
3. Respond – Consent Required (Respond to Webhook): sends the response back to the webhook caller
4. Check – File Type (PDF/Image) (If): checks a condition and branches into true / false
5. Upload – JigsawStack (Storage) (HTTP Request): calls api.jigsawstack.com (POST)
6. Extract – PDF Metadata/Text (Extract from File): extracts data from a file (pdf)
7. Classify – Image Files (HTTP Request): calls api.jigsawstack.com (POST)
8. Check – File Size & Pages (If): checks a condition and branches into true / false
9. Respond – Classification Result (Image) (Respond to Webhook): sends the response back to the webhook caller
10. Analyze – Confidence Score (Heuristics) (Code): runs custom JavaScript code

…plus 5 more steps. See the full node list below.

HTTP calls to: api.jigsawstack.com

Before you upgrade to n8n 3.0

Nodes used (8)

• Ifn8n-nodes-base.if
• Webhookn8n-nodes-base.webhook
• Respond to Webhookn8n-nodes-base.respondToWebhook
• Coden8n-nodes-base.code
• Extract from Filen8n-nodes-base.extractFromFile
• Switchn8n-nodes-base.switch
• HTTP Requestn8n-nodes-base.httpRequest
• No Operation, do nothingn8n-nodes-base.noOp

From the author

<section <h2🌊 What it Does</h2 <p This workflow <strongautomatically classifies uploaded files</strong (PDFs or images) as <spanfloorplans</span or <spannon‑floorplans</span. It filters out junk files, then analyzes valid floorplans to extract <emroom…

View the original on n8n.io → Check your own copy