How email-in extraction works
DocParse gives every extraction a unique inbound email address (enable Email Import on the extraction's configuration page). Anything sent to that address is processed automatically: each attachment is picked up, run through your extraction schema, and the results appear in the batch like any uploaded file — same fields, same validation, same exports.
- One address per workflow — your invoices address, your delivery-notes address, each with its own schema
- Multiple attachments per email are all processed (PDF, PNG, JPG, WEBP, DOCX)
- Results flow through the same pipeline: validation rules, review queue, exports, webhooks
Three setups, from manual to invisible
Forwarding is the zero-infrastructure start: anyone on the team forwards a supplier email to the extraction address and the data is in the system before they've closed the message.
Inbox rules make it automatic: a Gmail/Outlook filter that forwards messages matching from:supplier with attachments to the address. Documents now extract themselves; your team only sees the review queue.
Supplier-direct removes you entirely: give high-volume senders the address as the official destination ('email invoices to inv-yourcompany@inbound...'). The inbox step disappears.
What happens after extraction
Emailed documents go through exactly the pipeline you'd want for risky input: validation rules check required fields and ranges, suspect documents land in the review queue with the original file alongside the extracted values, and confirmed data leaves via Excel/CSV export, signed webhooks to your endpoint, or Zapier into your accounting tool, sheet or database.
That review step matters more for email than for any other channel — you don't control what people send. A blurry photo of a crumpled receipt gets flagged, not silently mis-posted.
A concrete example: supplier invoices
Create an extraction from the invoice template, enable Email Import, and add a Gmail filter forwarding supplier emails to the address. From that point: supplier sends invoice → extraction runs on arrival → validation requires invoice number and total → clean invoices auto-export to your webhook; odd ones wait in review. The human role shrinks from data entry to a daily two-minute review pass.
Frequently asked questions
Can I extract data from email bodies, not just attachments?
DocParse's email import processes attachments — the documents themselves. If your data lives in the email body text, save the email as a file and upload it, or use the API.
Is emailing documents to a processing address secure?
Each address is unique and unguessable per extraction. Documents are encrypted in transit, processed under the same rules as uploads (no training on your data), and deletable any time. Details at docparse.in/security.
What file types work as email attachments?
PDF, PNG, JPG, WEBP and DOCX attachments are processed, up to 25 MB per file — the same limits as dashboard uploads.