Legal & Advice

Claim File Extraction Agent

Before anyone can compare the figures, they have already typed them in twice.

Extracts the figures from expert reports and settlement letters and sets them side by side.

Extracts the figures from expert reports and settlement letters and sets them side by side.

Open agent

Enlarge demo

Public administration

Four processes that take up the most time today

Developed in Germany · Hosted in the EU

Developed in Germany · Hosted in the EU or on-premises

Delivered together with

Sovereign hosting in Germany, single-tenant from the first user.

Sovereign hosting in Germany, single-tenant from the first user.

Legal, tax and compliance expertise for legally sound deployment.

Legal, tax and compliance expertise for legally sound deployment.

Advisory and implementation for the public sector, starting with the AI readiness assessment.

Advisory and implementation for the public sector, starting with the AI readiness assessment.

An expert report with 42 line items is re-entered in full for the comparison.

The expert report arrives as a scan from the assessor; the insurer's settlement letter is a PDF in the inbox. Each lists its line items in its own order and under its own labels. Anyone who wants to set them side by side first builds a spreadsheet and enters every amount by hand, once when the file is opened and a second time when the assessment is due.

42

report line items

The assessor lists spare parts, labour units and paintwork individually.

31

settled lines

The settlement letter has fewer lines and labels them differently.

73

amounts typed by hand

For the comparison, every amount from both documents is typed into a spreadsheet.

11

lines without a match

Eleven report items do not appear in the settlement at all and have to be hunted down one by one for the comparison.

The 73 manual entries disappear and the eleven unmatched lines are already flagged in the line item comparison, which cuts data entry by 60% before the case handler looks at the first difference.

The 60% reduction in data entry is this agent's headline figure. The line item counts and manual entries describe a typical property damage case at a law firm; none of it was measured.

Three steps through the process.

The agent extracts the expert report and the settlement once and sets every line item side by side with amount and origin; none of the lines passes judgement on a deduction.

1.

1.

Extract the expert report and the settlement

The agent reads both documents page by page and captures every line item with label, amount and source reference. It flags lines from poor scans rather than guessing the amount.

The agent reads both documents page by page and captures every line item with label, amount and source reference. It flags lines from poor scans rather than guessing the amount.

File SR-2211/26, 42 report line items and 31 settlement lines captured

2.

2.

Match the line items to each other

The agent matches the same services in both documents, even where they are labelled differently. Lines without a match stay as separate lines.

The agent matches the same services in both documents, even where they are labelled differently. Lines without a match stay as separate lines.

Line item comparison with 31 pairs and 11 lines without a match

3.

3.

File the line item comparison in the case file

The agent files the table with amount, origin and page reference in the case file. The differences are then assessed on that basis.

The agent files the table with amount, origin and page reference in the case file. The differences are then assessed on that basis.

positionsspiegel_sr-2211-26.xlsx, 42 lines with page reference

What the agent does not do.

It does not assess the differences.

It does not assess the differences.

It sets them side by side with amount and source reference. The claims case handler decides whether a deduction is accepted or challenged.

It does not correct any amount in the expert report.

It does not correct any amount in the expert report.

If it cannot read a line in the scan reliably, it flags it in the line item comparison with page and excerpt. The secretariat re-enters it.

It does not capture handwritten notes in the margins of the report.

It does not capture handwritten notes in the margins of the report.

The note “blending disputed” next to item 27 is not in the line item comparison. Anyone who needs it will find it in the scan on page 6.

Connected systems

Practice management software

document management

scanned mail inbox

Operations

Deployment model

EU cloud, private cloud or on-premises

EU cloud, private cloud or on-premises

Data processing

GDPR-compliant, no model training on your data

GDPR-compliant, no model training on your data

Responsible teams

Law firms, claims case handlers

Law firms, claims case handlers

The questions that come up before the first meeting.

How quickly is the Claim File Extraction Agent ready to use?

Where is the data processed?

Where does the figure “60% less data entry” come from?

Related agents.

All agents

Platform

Industries

Services

Resources

Company

Pricing

From AI strategy to a running platform: we show you the way.

Talk to us

Log in

DE

EN

English

Platform

Industries

Services

Resources

Company

Pricing

From AI strategy to a running platform: we show you the way.

Talk to us

Log in

DE

EN

English
neuland.ai

The AI management and orchestration platform from Cologne.

Developed in Cologne. Operated in the EU or in your data center.

Certifications and compliance: ISO 27001, ISO 9001, GDPR, EU AI Act, DORA, StBerG, and BRAO
WirtschaftsWoche Best of Technology 2026 seal, 2nd place in the Process Management category for neuland.ai

© 2026 neuland AI AG