Legacy Contract Import
When transitioning from another system to Agiloft, or when receiving any large batch of contract documents, you can import and parse all those documents in Agiloft in a few easy steps.
When you import a batch of documents, Agiloft uses AI to analyze each document, pull out the most important information, and apply that data to discrete fields in individual Contract records. The extraction process is configurable, and after the AI analysis is complete, you can review the analysis to make sure the extracted data is accurate.
Prerequisites
To perform Legacy Import projects with AI, you need to have an AI Platform license and have AI Platform deployed.
Importing Contracts
To begin a Legacy Contract Import project:
- In your KB, navigate to the AI Projects table.
- Create an AI Project.
- In the Type field, choose Legacy Import.
- Name the project.
- Click Upload Documents. This also saves the record.
- Click Add Documents.
- Add each contract you'd like to process. As each document is added, a Document Analysis action tags key terms and clauses, and marks the Status as Ready to Review.
- For each document:
- Open the document from the Documents embedded table in the AI Projects record.
- Key terms are highlighted in yellow, clauses are highlighted in blue. Information about the tagged labels is shown in the pane on the left-hand side. Click the first label shown on the pane.
- Click Accept or Remove based on whether the model has tagged the proper information. You can also Edit from the options menu.
- Repeat for all labels in the pane.
- Check the document for any key terms or clauses that AI may have missed. You can tag missing information by highlighting it, selecting the proper tag, and filling out the form that appears in the left pane.
- When you're satisfied with the tags, click Mark Reviewed in the top right-hand corner to update the document status to Reviewed.
- Once all your documents are Reviewed, select them and click Publish. This puts each document in a state where their information can be pushed into a Contract record.
- In the Key Terms section, map each extracted label to a field in the new Contract record.
- Choose the criteria that each key term label requires before it can be published. You can choose any or all of the available options for each label:
- Select "AI extractions pending review" to publish labels that were tagged automatically through Add Documents in step 6.
- Select "Accepted AI Extractions" to publish labels that were approved manually in step 6c.
- Select "Manually added tags" to publish labels that were tagged manually.
- In the Clauses section, you can usually leave the Field Mapping section's values as Not Mapped. This is only required if you plan on displaying full clauses in record fields.
- Choose the criteria that each clause label requires before it can be published. You can choose any number of the available options for each label:
- Select "AI extractions pending review" to publish labels that were tagged automatically through Add Documents in step 6.
- Select "Accepted AI Extractions" to publish labels that were approved manually in step 6c.
- Select "Manually added tags" to publish labels that were tagged manually.
- Click Publish. This creates:
- A new Contract record for each document selected in step 7, with the key terms populated in the record based on the selections in step 9.
- A new Attachment record for each document selected in step 7, which holds the actual document you uploaded.
- New Contract Clause records, with the clauses that were reviewed in step 11.
This completes the workflow.
Document Analysis Action
If you don't want to extract all key terms and clauses during a legacy import project, you can modify which labels are tagged by reconfiguring the Document Analysis action in your KB.