Skip to main content

Validating Extracted Contract Fields

How to validate extracted contract fields in the data validation panel.

Written by Antonio Goncalves

Review and confirm the data Vallor extracts from your contracts to keep your library accurate.

Role tags

Before starting

Note:
You need at least one contract uploaded to your Library. Vallor must have completed its initial analysis of the contract before fields appear in the Extracted Data tab.

Overview

What this flow does

When Vallor analyses a contract, it automatically extracts key fields: supplier name, contract value, start and end dates, payment terms, governing law, and more. Each extracted field can be validated (checkmark) or flagged as incorrect (X).

This trains the AI and ensures your contract library stays accurate over time.


Step 1

Step 1: Open the contract in the Library

Click Library in the left sidebar. Find the contract you want to review and click to open it.

app.vallor.ai/library

Step 2

Step 2: Go to the Extracted Data tab

In the contract view, click the Extracted Data tab. This shows all fields Vallor has extracted from the document.

Extracted Data tab in contract view

If the Extracted Data tab is empty or shows a loading state, the AI analysis may still be in progress. Wait a moment and refresh.


Step 3

Step 3: Review each extracted field

Vallor pulls out key fields automatically. Read each value and check it against the actual contract text.

Fields typically extracted include:

  • Supplier name

  • Contract value

  • Start date and end date

  • Payment terms

  • Governing law

  • Contract type

  • Notice period

  • Renewal terms

Tip: Pay special attention to date fields and monetary values. These are the most common extraction errors on complex contracts.

Reviewing extracted contract fields

Step 4

Step 4: Confirm a correct field

If the extracted value is accurate, click the checkmark () icon next to the field.

This confirms the extraction and marks it as validated. It also signals to the AI that this extraction pattern is correct, improving future extractions on similar contracts.

Validating an extracted field with AI confidence popup

Tip: Even if a field looks right, clicking checkmark is worth doing. It actively reinforces the AI model and reduces errors on similar contracts in the future.


Step 5

Step 5: Flag an incorrect field

If the value is wrong or missing, click the X icon next to the field.

A text input appears. Enter the correct value and save. The corrected value is stored and feeds back into the AI model to prevent the same error on future contracts.

Tip: If a field is genuinely not applicable (for example, no renewal clause exists in the contract), click checkmark with the value blank. This tells the AI the absence is correct, not a missed extraction.


Step 6

Step 6: Complete all fields before proceeding

Validate every field before using the contract for reporting or analysis. Unvalidated fields are shown with a neutral indicator and are excluded from accuracy scoring.


Tips section

Tips and workarounds

Tip: Even if a field looks right, clicking checkmark signals to the AI that this extraction pattern is correct. It improves future extractions on similar contracts.

Tip: Date fields and monetary values are the most common extraction errors on complex contracts. Double-check these against the original document text.

Tip: If a field is genuinely not applicable (e.g. no renewal clause exists), click checkmark with the value blank. This tells the AI the absence is correct, not a missed extraction.


What to do next

What to do next

  • Run AI Redlining on a Supplier Contract

  • Add Supporting Documents in Chat

  • Create a Negotiation Playbook


Did this answer your question?