ACORD extraction

ACORD and loss run data extraction with page citations.

Upload the ACORD 125, line supplements and loss runs. PacketSlip pulls out the fields underwriters care about and shows exactly where each value came from.

ACORD 125 · page 1 of 4
APPLICANT: Brightline Electric LLC · FEIN 84-•••••12
NAICS 238210 · Electrical contractors
Extracted fields
FieldValueSource
Named insuredBrightline Electric LLCp.1
FEIN84-•••••12p.1
NAICS238210p.1
Annual revenue$4,200,000p.2
Payroll$1,350,000p.2
Employees28p.2

What it does

Structured fields, not raw text

Named insured, FEIN, addresses, NAICS, revenue, payroll, employee count, prior carrier and loss totals land in a clean, editable list.

A citation for every value

Click any value to jump to the page it was read from. No guessing whether a number came from page 2 or page 9.

Multi-document packets

ACORD 125, 126, 130, 140, supplements, SOVs and carrier loss runs can be uploaded together and are read as one submission.

Preflight before reading

Encrypted, empty, oversized or image-only scans are caught at upload, so you know immediately if a file cannot be read.

How it works

  1. 1

    Drop in the packet

    Drag the PDFs onto the submission. Each file is checked for size, page count and readability.

  2. 2

    Fields are extracted

    Key values are read from each document and tagged with document name, page and confidence.

  3. 3

    Confirm and edit

    Review low-confidence values side by side with the source page and correct anything in one click.

Before and after

The manual way

  • Re-keying the same ACORD data into the agency system and carrier portals
  • Scrolling through 40-page packets to find one payroll figure
  • No way to show an underwriter where a number came from

With PacketSlip

  • Key fields extracted in minutes, ready to copy
  • One click from a value to its source page
  • Every value carries document, page and confidence

What extraction looks like on real packets

Retail agency

Renewal packet with five years of loss runs

Situation
A contractor renewal arrives as one 38-page PDF: ACORD 125, 126, and loss runs from two carriers.
What PacketSlip does
Each document type is detected, applicant and coverage fields are extracted, and loss run totals are captured per policy year with page references.
Outcome
The account manager reviews structured data instead of re-keying it, and clicks any value to see the original page.

Scanned forms

Partly handwritten application

Situation
An insured prints, fills out by hand and scans the application.
What PacketSlip does
Fields read with lower certainty are marked so a person checks them, instead of being silently accepted.
Outcome
Reviewers spend time only on the values that need attention.

Wholesale

Packet forwarded from a retail agent

Situation
A wholesaler receives a submission with an application, SOV spreadsheet export and a carrier loss run.
What PacketSlip does
Key risk details are pulled into one view so the submission can be triaged against appetite quickly.
Outcome
Decline-or-proceed decisions take minutes rather than a full read-through.

Documents and fields covered

DocumentExamples of extracted fields
ACORD 125 – Commercial Insurance ApplicationNamed insured, FEIN, mailing address, NAICS / SIC, business description, years in business, requested lines, effective dates
ACORD 126 – General LiabilityHazard classifications, exposure basis, revenue, subcontracted work, limits requested
ACORD 140 – PropertyLocations, construction, occupancy, building and contents values, protection class
ACORD 127 / 137 – AutoVehicle and driver counts, garaging addresses, radius of operation
ACORD 130 – Workers' CompensationClass codes, payroll per class, employee counts, experience modification
Carrier loss runsPolicy periods, claim counts, paid and reserved amounts, open claims

Which forms are supported

PacketSlip is built for commercial P&C packets: ACORD 125 (Commercial Insurance Application), ACORD 126 (General Liability), ACORD 130 (Workers' Compensation), ACORD 131 (Umbrella), ACORD 137 and 127 (Auto), ACORD 140 (Property), plus carrier supplements, statements of values and loss runs.

Fillable and flattened PDFs both work. Scanned documents are flagged so you know when a value needs a closer look.

Every value comes with evidence

Extraction is only useful if people trust it. Each value is stored with the document and page it came from and shown with a confidence level. Clicking a value jumps the document viewer to that page.

When a field cannot be found, it is reported as missing rather than guessed. Missing required fields flow into the checklist and the follow-up email.

Preflight before extraction

Before anything is read, uploaded files are checked: password protection, damaged files, empty pages and files above the size limit are caught immediately with a clear message, so problems surface while the uploader is still at their desk.

Frequently asked questions

Does it work on scanned ACORD forms?

Text-based PDFs give the best results. Image-only scans are flagged during preflight, and values read from them are marked for review.

Can I correct an extracted value?

Yes. Edits are saved with your name and time in the submission history, and the original value is kept.

Which fields are extracted?

Named insured, FEIN, mailing and location addresses, NAICS/SIC, business description, years in business, revenue, payroll, employee count, prior carrier, expiring premium and loss summary, among others.

Is our data used to train AI models?

No. Documents are stored privately for your agency and are only used to answer questions about that submission.

Do I need to label which document is which?

No. Document types are detected automatically. You can correct the type if needed.

What file formats are supported?

PDF files up to 20 MB each. Multiple documents can be uploaded to one submission.