Introducing s-1: Sievly’s new document reading model
All posts
Cookbooks
Cookbook: extract line items from invoices
A step-by-step recipe for pulling vendor, totals and line items from mixed invoice formats with a single Extract schema.
Priya Nair
·
Developer Advocate
sievly
Invoice line items
Invoices are a good first project: every company has them, and every vendor formats them differently. This recipe walks through a schema that holds up across formats.
1. Define the schema
Start with the fields you actually use: vendor name, invoice number, issue date, currency, total and a list of line items with description, quantity and amount.
2. Run Extract
Send the file to Extract with your schema. Each field comes back with a citation, so you can show reviewers exactly where a value came from.
3. Handle the exceptions
Route invoices with missing totals or low-confidence line items to a review step in Workbench instead of failing the whole batch.
4. Ship it
Save the schema as a pipeline in Workbench and call it by ID from your accounts payable system.
More from the blog
sievly
Introducing s-1
Announcements
Introducing s-1, our new document reading model
s-1 reads PDFs, scans and spreadsheets the way a careful analyst does, and returns structured data with a citation for every value. It is available in preview today.
sievly
Reading order
Research
Research note: reading order in complex layouts
Multi-column pages, sidebars and callouts confuse simple reading order. We share what we learned teaching s-1 to follow the page the way a person would.
sievly
Quarryline × Sievly
Case Studies
How Quarryline turns research notes into a searchable knowledge base
Quarryline’s data team uses Parse and Extract to turn thousands of analyst notes, filings and spreadsheets into clean, cited records their AI assistant can trust.