Resize, sharpen, and compress an image to fit email platform size limits in a single pipeline.
The English summary is being prepared. Showing the original description.
Extract requestor, ship-from, ship-to, service, package, and handling instructions from shipping requests.
The English summary is being prepared. Showing the original description.
Review the instructions the agent will receive before you install.
Operations teams use this recipe to extract internal shipping request fields into structured records for carrier booking, warehouse fulfillment, and label generation.
Document Extraction (1 credit per page)
You need an Iteration Layer API key. Get one at platform.iterationlayer.com during the 7-day trial.
For full integration guidance (SDKs, auth, MCP, error handling), see the Iteration Layer Integration Guide.
curl -X POST https://api.iterationlayer.com/document-extraction/v1/extract \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"files": [
{
"type": "url",
"name": "shipping-request.pdf",
"url": "https://example.com/documents/shipping-request-sample.pdf"
}
],
"schema": {
"fields": [
{
"name": "request_number",
"type": "TEXT",
"description": "Shipping request number"
},
{
"name": "requestor_name",
"type": "TEXT",
"description": "Requestor name"
},
{
"name": "department_or_cost_center",
"type": "TEXT",
"description": "Department, project, or cost center"
},
{
"name": "ship_from",
"type": "ADDRESS",
"description": "Ship-from address"
},
{
"name": "ship_to",
"type": "ADDRESS",
"description": "Ship-to address"
},
{
"name": "requested_ship_date",
"type": "DATE",
"description": "Requested ship date"
},
{
"name": "service_level",
"type": "TEXT",
"description": "Requested carrier service level"
},
{
"name": "packages",
"type": "ARRAY",
"description": "Packages requested for shipment",
"fields": [
{
"name": "description",
"type": "TEXT",
"description": "Package contents or description"
},
{
"name": "quantity",
"type": "DECIMAL",
"description": "Package quantity"
},
{
"name": "weight_kg",
"type": "DECIMAL",
"description": "Package weight in kilograms"
},
{
"name": "dimensions",
"type": "TEXT",
"description": "Package dimensions"
}
]
},
{
"name": "special_handling",
"type": "TEXT",
"description": "Fragile, dangerous goods, refrigeration, insurance, or other handling instructions"
}
]
}
}'
import { IterationLayer } from "iterationlayer";
const client = new IterationLayer({ apiKey: "YOUR_API_KEY" });
const result = await client.extractDocument({
"files": [
{
"type": "url",
"name": "shipping-request.pdf",
"url": "https://example.com/documents/shipping-request-sample.pdf"
}
],
"schema": {
"fields": [
{
"name": "request_number",
"type": "TEXT",
"description": "Shipping request number"
},
{
"name": "requestor_name",
"type": "TEXT",
"description": "Requestor name"
},
{
"name": "department_or_cost_center",
"type": "TEXT",
"description": "Department, project, or cost center"
},
{
"name": "ship_from",
"type": "ADDRESS",
"description": "Ship-from address"
},
{
"name": "ship_to",
"type": "ADDRESS",
"description": "Ship-to address"
},
{
"name": "requested_ship_date",
"type": "DATE",
"description": "Requested ship date"
},
{
"name": "service_level",
"type": "TEXT",
"description": "Requested carrier service level"
},
{
"name": "packages",
"type": "ARRAY",
"description": "Packages requested for shipment",
"fields": [
{
"name": "description",
"type": "TEXT",
"description": "Package contents or description"
},
{
"name": "quantity",
"type": "DECIMAL",
"description": "Package quantity"
},
{
"name": "weight_kg",
"type": "DECIMAL",
"description": "Package weight in kilograms"
},
{
"name": "dimensions",
"type": "TEXT",
"description": "Package dimensions"
}
]
},
{
"name": "special_handling",
"type": "TEXT",
"description": "Fragile, dangerous goods, refrigeration, insurance, or other handling instructions"
}
]
}
});
from iterationlayer import IterationLayer
client = IterationLayer(api_key="YOUR_API_KEY")
result = client.extract_document(**{
"files": [
{
"type": "url",
"name": "shipping-request.pdf",
"url": "https://example.com/documents/shipping-request-sample.pdf"
}
],
"schema": {
"fields": [
{
"name": "request_number",
"type": "TEXT",
"description": "Shipping request number"
},
{
"name": "requestor_name",
"type": "TEXT",
"description": "Requestor name"
},
{
"name": "department_or_cost_center",
"type": "TEXT",
"description": "Department, project, or cost center"
},
{
"name": "ship_from",
"type": "ADDRESS",
"description": "Ship-from address"
},
{
"name": "ship_to",
"type": "ADDRESS",
"description": "Ship-to address"
},
{
"name": "requested_ship_date",
"type": "DATE",
"description": "Requested ship date"
},
{
"name": "service_level",
"type": "TEXT",
"description": "Requested carrier service level"
},
{
"name": "packages",
"type": "ARRAY",
"description": "Packages requested for shipment",
"fields": [
{
"name": "description",
"type": "TEXT",
"description": "Package contents or description"
},
{
"name": "quantity",
"type": "DECIMAL",
"description": "Package quantity"
},
{
"name": "weight_kg",
"type": "DECIMAL",
"description": "Package weight in kilograms"
},
{
"name": "dimensions",
"type": "TEXT",
"description": "Package dimensions"
}
]
},
{
"name": "special_handling",
"type": "TEXT",
"description": "Fragile, dangerous goods, refrigeration, insurance, or other handling instructions"
}
]
}
})
package main
import il "github.com/iterationlayer/sdk-go"
func main() {
client := il.NewClient("YOUR_API_KEY")
result, err := client.ExtractDocument(il.ExtractDocumentRequest{
Files: []il.FileInput{
il.FileInput{
Type: "url",
Name: "shipping-request.pdf",
Url: "https://example.com/documents/shipping-request-sample.pdf",
},
},
Schema: il.ExtractionSchema{
Fields: []any{
il.TextFieldConfig{
Name: "request_number",
Type: "TEXT",
Description: "Shipping request number",
},
il.TextFieldConfig{
Name: "requestor_name",
Type: "TEXT",
Description: "Requestor name",
},
il.TextFieldConfig{
Name: "department_or_cost_center",
Type: "TEXT",
Description: "Department, project, or cost center",
},
il.AddressFieldConfig{
Name: "ship_from",
Type: "ADDRESS",
Description: "Ship-from address",
},
il.AddressFieldConfig{
Name: "ship_to",
Type: "ADDRESS",
Description: "Ship-to address",
},
il.DateFieldConfig{
Name: "requested_ship_date",
Type: "DATE",
Description: "Requested ship date",
},
il.TextFieldConfig{
Name: "service_level",
Type: "TEXT",
Description: "Requested carrier service level",
},
il.ArrayFieldConfig{
Name: "packages",
Type: "ARRAY",
Description: "Packages requested for shipment",
Fields: []any{
il.TextFieldConfig{
Name: "description",
Type: "TEXT",
Description: "Package contents or description",
},
il.DecimalFieldConfig{
Name: "quantity",
Type: "DECIMAL",
Description: "Package quantity",
},
il.DecimalFieldConfig{
Name: "weight_kg",
Type: "DECIMAL",
Description: "Package weight in kilograms",
},
il.TextFieldConfig{
Name: "dimensions",
Type: "TEXT",
Description: "Package dimensions",
},
},
},
il.TextFieldConfig{
Name: "special_handling",
Type: "TEXT",
Description: "Fragile, dangerous goods, refrigeration, insurance, or other handling instructions",
},
},
},
})
if err != nil {
panic(err)
}
_ = result
}
{
"name": "Extract Shipping Request Data",
"nodes": [
{
"parameters": {
"content": "## Extract Shipping Request Data\n\nOperations teams use this recipe to extract internal shipping request fields into structured records for carrier booking, warehouse fulfillment, and label generation.\n\n**Note:** This workflow uses the Iteration Layer community node (`n8n-nodes-iterationlayer`). Install it via Settings > Community Nodes on self-hosted n8n, or add it directly on n8n Cloud with Verified Community Nodes enabled.",
"height": 280,
"width": 500,
"color": 2
},
"type": "n8n-nodes-base.stickyNote",
"typeVersion": 1,
"position": [
200,
40
],
"id": "extract-shipping-request-data-overview",
"name": "Overview"
},
{
"parameters": {},
"type": "n8n-nodes-base.manualTrigger",
"typeVersion": 1,
"position": [
250,
300
],
"id": "extract-shipping-request-data-trigger",
"name": "Manual Trigger"
},
{
"parameters": {
"resource": "documentExtraction",
"schemaInputMode": "rawJson",
"schemaJson": "{\n \"fields\": [\n {\n \"name\": \"request_number\",\n \"type\": \"TEXT\",\n \"description\": \"Shipping request number\"\n },\n {\n \"name\": \"requestor_name\",\n \"type\": \"TEXT\",\n \"description\": \"Requestor name\"\n },\n {\n \"name\": \"department_or_cost_center\",\n \"type\": \"TEXT\",\n \"description\": \"Department, project, or cost center\"\n },\n {\n \"name\": \"ship_from\",\n \"type\": \"ADDRESS\",\n \"description\": \"Ship-from address\"\n },\n {\n \"name\": \"ship_to\",\n \"type\": \"ADDRESS\",\n \"description\": \"Ship-to address\"\n },\n {\n \"name\": \"requested_ship_date\",\n \"type\": \"DATE\",\n \"description\": \"Requested ship date\"\n },\n {\n \"name\": \"service_level\",\n \"type\": \"TEXT\",\n \"description\": \"Requested carrier service level\"\n },\n {\n \"name\": \"packages\",\n \"type\": \"ARRAY\",\n \"description\": \"Packages requested for shipment\",\n \"fields\": [\n {\n \"name\": \"description\",\n \"type\": \"TEXT\",\n \"description\": \"Package contents or description\"\n },\n {\n \"name\": \"quantity\",\n \"type\": \"DECIMAL\",\n \"description\": \"Package quantity\"\n },\n {\n \"name\": \"weight_kg\",\n \"type\": \"DECIMAL\",\n \"description\": \"Package weight in kilograms\"\n },\n {\n \"name\": \"dimensions\",\n \"type\": \"TEXT\",\n \"description\": \"Package dimensions\"\n }\n ]\n },\n {\n \"name\": \"special_handling\",\n \"type\": \"TEXT\",\n \"description\": \"Fragile, dangerous goods, refrigeration, insurance, or other handling instructions\"\n }\n ]\n}",
"files": {
"fileValues": [
{
"fileInputMode": "url",
"fileName": "shipping-request.pdf",
"fileUrl": "https://example.com/documents/shipping-request-sample.pdf"
}
]
}
},
"type": "n8n-nodes-iterationlayer.iterationLayer",
"typeVersion": 1,
"position": [
500,
300
],
"id": "extract-shipping-request-data-extract",
"name": "Extract Data",
"credentials": {
"iterationLayerApi": {
"id": "1",
"name": "Iteration Layer API"
}
}
}
],
"connections": {
"Manual Trigger": {
"main": [
[
{
"node": "Extract Data",
"type": "main",
"index": 0
}
]
]
}
},
"settings": {
"executionOrder": "v1"
}
}
Extract shipping request data from the file at [file URL]. Use the extract_document tool with these fields:
- request_number (TEXT): Shipping request number
- requestor_name (TEXT): Requestor name
- department_or_cost_center (TEXT): Department, project, or cost center
- ship_from (ADDRESS): Ship-from address
- ship_to (ADDRESS): Ship-to address
- requested_ship_date (DATE): Requested ship date
- service_level (TEXT): Requested carrier service level
- packages (ARRAY): Each with description (TEXT), quantity (DECIMAL), weight_kg (DECIMAL), dimensions (TEXT)
- special_handling (TEXT): Fragile, dangerous goods, refrigeration, insurance, or other handling instructions
{
"success": true,
"data": {
"request_number": {
"value": "Request Number",
"confidence": 0.97,
"citations": [
"REQUEST NUMBER"
]
},
"requestor_name": {
"value": "Requestor Name",
"confidence": 0.97,
"citations": [
"REQUESTOR NAME"
]
},
"department_or_cost_center": {
"value": "Department Or Cost Center",
"confidence": 0.97,
"citations": [
"DEPARTMENT OR COST CENTER"
]
},
"ship_from": {
"value": {
"street": "100 Market Street",
"city": "Berlin",
"postal_code": "10115",
"country": "DE"
},
"confidence": 0.97,
"citations": [
"SHIP FROM"
]
},
"ship_to": {
"value": {
"street": "100 Market Street",
"city": "Berlin",
"postal_code": "10115",
"country": "DE"
},
"confidence": 0.97,
"citations": [
"SHIP TO"
]
},
"requested_ship_date": {
"value": "2026-04-15",
"confidence": 0.97,
"citations": [
"15 Apr 2026"
]
},
"service_level": {
"value": "Service Level",
"confidence": 0.97,
"citations": [
"SERVICE LEVEL"
]
},
"packages": {
"value": [
{
"description": {
"value": "Description",
"confidence": 0.96,
"citations": [
"DESCRIPTION"
]
},
"quantity": {
"value": 12.5,
"confidence": 0.96,
"citations": [
"QUANTITY"
]
},
"weight_kg": {
"value": 12.5,
"confidence": 0.96,
"citations": [
"WEIGHT KG"
]
},
"dimensions": {
"value": "Dimensions",
"confidence": 0.96,
"citations": [
"DIMENSIONS"
]
}
}
],
"confidence": 0.97,
"citations": [
"PACKAGES"
]
},
"special_handling": {
"value": "Special Handling",
"confidence": 0.97,
"citations": [
"SPECIAL HANDLING"
]
}
}
}
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Resize, sharpen, and compress an image to fit email platform size limits in a single pipeline.
The English summary is being prepared. Showing the original description.
Compress an image to fit within a specific file size in bytes using quality-first compression.
The English summary is being prepared. Showing the original description.
Convert a contract PDF to clean markdown for clause extraction or LLM analysis.
The English summary is being prepared. Showing the original description.
Convert external documents — specs, contracts, reports — to markdown for knowledge base ingestion.
The English summary is being prepared. Showing the original description.
Convert a document to clean markdown suitable for chunking and embedding in a RAG pipeline.
The English summary is being prepared. Showing the original description.
Convert an image between PNG, JPEG, and WebP formats with quality control for web optimization.
The English summary is being prepared. Showing the original description.