Introducing s-1: Sievly’s new document reading model
Split
Break long files into the documents inside them
Find the boundaries in multi-document files so every part reaches the right step of your pipeline.
POST /split
Helping everyone from early-stage startups to global enterprises unlock their data.
Quarryline
Brindlewood
nimbleton
brightwick
vantrel
PLINTH & CO
OAKMERE
ARROWFEN
Halvory
pebblegate
Mirrowind
quillstack
Split
One packet, many documents
DEFINITION
Split looks at a long file, finds where one document ends and the next begins, and returns page ranges with a label for each part.
WHO IT’S FOR
Teams that receive packets: claims, loan files, onboarding bundles, discovery sets and medical records.
THE PROBLEM IT SOLVES
Running extraction on a whole packet mixes fields from different documents. Splitting first keeps every part clean.
Split in the platform
Using Split in your pipeline
Split is the step before extraction when one file holds many documents.
#1
When files arrive as packets
Break claim packets, loan files and onboarding bundles into their parts.
#2
When each part needs its own schema
Route each section to the right Extract schema.
#3
When order matters
Keep page order and the links between parts.
Where teams use Split
Packets that used to be sorted by hand
Common jobs where Split does the sorting.
Claim packets
Separate notices, estimates, invoices and letters.
Loan files
Find statements, pay slips and IDs inside one upload.
Discovery sets
Split productions into individual documents.
Medical records
Separate notes, results and referrals.
Onboarding bundles
Break employee or customer packets into forms.
Bid packages
Divide specifications, drawings and forms.
Split a packet of your own in Workbench.
Why Split
Why teams use Split
01
Page-level boundaries
Each part comes back as a clear page range.
02
Labels you define
Name the document types that matter to you.
03
Works with Classify
Label every part before extraction.
04
Built for long packets
Handle files with many similar documents.
How Split works
How Split works in four steps
■ STEP 01
Define your types
List the documents you expect in a packet.
types: […]
■ STEP 02
Send the packet
Call Split with the file.
client.split(file)
■ STEP 03
Inspect the parts
See each range beside the source pages.
pages 1 to 4
■ STEP 04
Route each part
Send parts on to Extract or storage.
→ extract
Built for production
Built for production workloads
Split runs on the same infrastructure as every Sievly endpoint, with the controls enterprise teams expect.
Built for regulated teams
Security controls cover the whole platform, with agreements available on Growth and Enterprise.
Deploy in your environment
Run Sievly in your own cloud when data cannot leave your network.
High-availability uptime
Built for steady throughput at any volume, with SLAs on Enterprise.
The rest of the platform
The endpoints around Split
/parse
Parse
Turn any document into structured, cited content.
Read more →
/extract
Extract
Pull the fields you define, with a citation on every value.
Read more →
/classify
Classify
Label and route documents by type before they move on.
Read more →
/edit
Edit
Fill forms and update documents from plain language.
Read more →
/workbench
Workbench
Design, test and deploy pipelines without code.
Read more →
Further reading