Skip to main content

Documentation Index

Fetch the complete documentation index at: https://echo-docs.beau.to/llms.txt

Use this file to discover all available pages before exploring further.

Echo extracts structured data from your emails and PDF attachments. Anything written in plain text can be picked up — if it’s in the email body or an attached PDF, Echo can read it.

​
What Echo extracts

Echo pulls out the types of data that naturally appear in emails: things like amounts, dates, names, and statuses. More specifically, Echo works with data that maps to standard schema.org types — a widely used vocabulary for describing real-world information. In practice, this means Echo can extract:
  • Amounts & prices — invoice totals, subscription fees, refund amounts
  • Dates & deadlines — due dates, expiry dates, flight times, tax deadlines
  • Names & organizations — vendors, senders, airlines, tax authorities
  • Reference numbers — invoice numbers, booking codes, PNR, tracking IDs
  • Statuses — paid, overdue, confirmed, delayed, cancelled
  • Promo codes & offers — discount codes, expiry, savings value
  • Addresses — billing addresses, delivery addresses
  • URLs & links — payment links, tracking pages
  • and more
If a value appears as readable text in an email, there’s a good chance Echo can extract it.

​
PDF attachments

Echo reads text from PDF attachments automatically, no setup needed. This includes invoices, receipts, boarding passes, and tax notices sent as attached files.
Images and scanned documents are not supported. Echo can only extract text it can read directly, not content embedded in photos or non-searchable PDFs.

​
What Echo doesn’t extract

  • Images — logos, photos, banners, and image-based content are ignored
  • Scanned PDFs — if the PDF is a photo of a document (non-searchable), Echo can’t read it
  • Data outside the email — Echo only reads what’s in the email body and its attachments, nothing else

​
A note on accuracy

Echo uses AI to interpret context, not just copy text. This means it can infer a status (“this invoice is overdue”) or normalize a format (“€1.200,00” → “$1,200.00”) — but it can also make mistakes. Always verify critical data against the original email.