Zapier and Make handle most automation use cases well. But when your destination is a custom application, an internal database, or a system that isn't in any app marketplace, you need a direct connection. That's what Parseur webhooks are for.
Every time Parseur processes a document, it sends an HTTP POST request to a URL you control, with the extracted fields as a JSON payload. Your application receives clean, structured data in real time and handles it however your system requires. No intermediary platform, no mapping through a third-party interface, no per-task billing.
A 2025 Parseur and QuestionPro survey found that employees spend more than 9 hours per week moving data out of documents and into other tools, at an average cost of $28,500 per employee per year. For development teams building document automation into their own products, webhooks are the lowest-latency path from parsed document to live system.
Key Takeaways
Parseur sends structured JSON to any HTTP endpoint via webhook every time a document is processed. No intermediary platform needed.
Four event types are supported: Document Processed, Flattened Tables, Table Item Processed, and Export Failed. Each targets a different use case.
Custom HTTP headers let you add authentication tokens or shared secrets to secure your webhook endpoint.
Webhooks work for any destination: custom apps, internal databases, ERPs, self-hosted automation tools like n8n, and microservices.
Use Zapier or Make when your destination is a mainstream app. Use webhooks when it isn't.
Webhook vs Zapier vs Make: when to use each
All three are ways to send Parseur data to other systems. The right choice depends on your destination:
Use Zapier or Make when your destination is a mainstream application already in their app libraries (Google Sheets, HubSpot, Salesforce, Slack, Airtable, and thousands more). Setup is faster and no coding is required.
Use a webhook when your destination is a custom application, proprietary internal system, a self-hosted tool like n8n, or a database that no-code platforms don't cover. Webhooks give you full control: you write the receiving endpoint, you decide what happens to the data, and you're not bound by any platform's data model or rate limits.
For teams building document automation into their own products or internal tools, webhooks are the standard integration path.
Webhook events Parseur supports
Parseur fires different events depending on what you need:
Event
When it fires
Best for
Document Processed
When a document is fully extracted
General-purpose: one payload per document with all fields
Flattened Tables
When a document containing tables is processed
Spreadsheet-style output where table rows are flattened
Table Item Processed
Once per row in a table field
Inserting each line item as a separate database record
Export Failed
When a webhook delivery attempt fails
Error monitoring and Slack/email alerts
The "Table Item Processed" event is especially useful for invoice line items or order tables, where you want each row to create a separate record rather than processing the whole array at once.
What the JSON payload contains
When a document is processed, Parseur sends a JSON object to your endpoint. The payload includes:
Your extracted fields as key-value pairs (field names matching the template you defined)
Table rows as arrays of objects, one object per row
Document metadata: document ID, mailbox ID, processing timestamp, and original filename
The exact structure mirrors your template. A field named invoice_number in your Parseur template arrives as invoice_number in the JSON. Table fields arrive as arrays, with each row containing the column values you defined.
Step 1: Upload a document and define your extraction template
Drag and drop a document into your mailbox, or forward an email to your Parseur mailbox address. Set up an auto-forwarding rule to handle incoming documents automatically at scale.
Once your document appears in Parseur, highlight the fields you want to extract and name each one. Parseur's AI applies that template to every future document of the same type.
Define the fields you want in your JSON payload
Step 2: Create your webhook
Go to Export, then Webhook, then New Webhook. Paste in your endpoint URL and choose your trigger event.
If you're testing the integration first, use webhook.site to get a temporary public URL that logs every incoming request. This lets you inspect the exact JSON payload before connecting your production endpoint.
Use webhook.site to inspect the payload before connecting your real endpoint
Paste the URL into Parseur under the target URL field.
Add your endpoint URL and choose the trigger event
Step 3: Add authentication headers (recommended)
For production endpoints, add a custom HTTP header with an authorization token or shared secret. Your receiving endpoint validates the header before processing the payload. This prevents unauthorized requests from reaching your system.
Step 5: Test the integration
Reprocess a document in your Parseur mailbox to trigger a webhook delivery. Check your endpoint logs or webhook.site to confirm the payload arrived and matches the structure you expect.
Verify the JSON payload structure in your endpoint logs
Once the test passes, your endpoint will receive a JSON payload automatically for every document Parseur processes.
Teams use Parseur webhooks to connect document parsing to systems that no-code platforms don't reach:
Custom databases: Push invoice line items directly into PostgreSQL or MySQL. Each "Table Item Processed" event inserts one row, keeping your schema clean without batch imports.
Internal ERPs: Route extracted purchase order data directly into an ERP system via its REST API, without building a file import pipeline.
Microservices: Trigger a downstream service when a document is processed, for example kicking off a validation step or notifying a fulfillment service when a new order arrives.
Self-hosted n8n: Use Parseur as the document extraction layer in an n8n workflow running on your own infrastructure. See the n8n integration guide for the full setup.
CRM and lead routing: Forward lead email data directly into a CRM's inbound API endpoint, bypassing Zapier entirely.
Free tools for related document types
If you need to inspect the JSON output from specific document types before building your integration, these free converters show you the structure in the browser:
Parseur is a powerful document processing software to extract text from emails, PDFs and documents and automate your workflow.
All Parseur features.
What is Webhook?
A webhook is used to exchange data between applications and servers using their API. A webhook is a notification event (via HTTP POST) that gets triggered every time a new document is parsed by Parseur and will send the parsed data as JSON.
Common questions about using Parseur webhooks to send parsed document data to custom applications and APIs.
A Parseur webhook is an HTTP POST request that Parseur sends to a URL you specify every time a document is processed. The request contains the extracted fields as a JSON payload. Your application receives the data in real time and can process, store, or route it however you need.
The JSON payload contains the extracted fields from your document as key-value pairs, along with metadata like the document ID, mailbox ID, and processing timestamp. Table data is included as an array of objects, with each row as a separate entry. The exact structure depends on your template fields. See the full payload reference in the Parseur documentation.
Parseur retries failed webhook deliveries automatically. If your endpoint is temporarily unavailable, the delivery will be retried. You can also use the Export Failed event to trigger an alert when a delivery attempt fails, so your team can respond quickly.
Yes. Self-hosted n8n instances connect to Parseur via generic webhook nodes. Your n8n instance needs to be publicly accessible so Parseur can reach the webhook endpoint. The Parseur native connector is also available for n8n Cloud.
Yes. The "Table Item Processed" event fires a separate webhook call for each row in a table field. This is useful when you want to insert each line item from an invoice as a separate database record, rather than receiving all rows in a single array.
Parseur supports four webhook trigger events. "Document Processed" fires with the full JSON payload when a document is extracted. "Flattened Tables" sends table data in a format where each table row is a separate object. "Table Item Processed" fires separately for each row in a table field. "Export Failed" fires when a delivery attempt fails, useful for error monitoring and alerts.
Yes. Parseur lets you add custom HTTP headers to your webhook request, including an Authorization header with a token or a shared secret. Your receiving endpoint can then validate each incoming request before processing it. This is the recommended approach for webhooks that write to production databases or trigger sensitive workflows.
Use a webhook when you're connecting to a custom application, proprietary internal system, or database that isn't available in Zapier or Make's app libraries. Webhooks give you full control over how the data is received and handled. If your destination is a mainstream app like Google Sheets, HubSpot, or Slack, Zapier or Make are faster to set up.
Yes. Tools like webhook.site give you a temporary public URL that logs every incoming request, so you can inspect the JSON payload structure before wiring it to your actual system. Reprocess a document in Parseur after setting up the webhook to trigger a test delivery.
Yes. Parseur has a REST API that lets you upload documents programmatically, retrieve parsed results, and manage mailboxes. Webhooks are push-based (Parseur calls your endpoint) while the API is pull-based (you call Parseur's endpoint). Most real-time integrations use webhooks; the API is useful for batch workflows or when you want to poll for results.
