In cases where the AI fails to correctly assign a paragraph to the appropriate clause, you have the ability to manually annotate the text. By tagging the paragraph yourself, you help improve the AI's performance, ensuring it gets better at clause identification and classification over time. This process allows you to either select an existing clause or create a new clause name for the paragraph, ensuring your annotations are accurately captured.
Steps to Annotate Text
Click on the Document Name to Open the Co-Pilot View
Start by navigating to the document you want to annotate. Click on the Document Name to open the Co-Pilot view, where you can interact with the document and begin the annotation process.
Scroll to the Text Not Predicted by AI
Scroll through the document to locate the text or paragraph that was not correctly predicted or assigned by the AI. The AI may miss certain clauses or obligations, so it’s essential to manually review and correct these instances.
Select the Text and Click on "Save as Clause"
Once you’ve found the text that needs to be tagged, highlight the specific paragraph or clause. After selecting the text, click on the Save as Clause button to initiate the tagging process.
Select the Clause Name or Create a New One
You will then have two options:
Select an Existing Clause: Choose a clause name from the predefined list of available clauses that best fits the selected text.
Enter a New Clause Name: If the existing clauses do not cover the selected text, you can select Other and manually enter a new clause name. This allows you to create custom clauses tailored to the content of the document.
Save the Annotation
Once you’ve chosen or created a clause name, click Save to store the annotation. This action will help train the AI for future assignments and ensure the text is classified correctly.

What’s Next?
After saving your annotation, the AI will learn from your manual input and improve its clause prediction accuracy over time. As you continue annotating, you will contribute to the AI’s ability to identify and classify clauses more effectively, streamlining the lease abstraction process.
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article