Now that we have the Issue's foundation, let's tell the AI what we need it to do. 

Click on the Issue, and in the far right column, click the tab for "AI Library."

This is where you can find the Concepts and Details for the AI and Document Searches. Concepts are legal concepts, such as Liability, Term, Governing Law, Force Majeure, etc., that the AI is trained to identify anywhere in a document. A detail is a specific piece of information like the length of a term (e.g., "two years") or the state of governing law (e.g., "Delaware").

• Concepts will list all of the Concepts in the skillset you selected. Hover over the one you need and click "Create Rule."
• The same applies to Details. Click "Create Rule" to add it to the Issue.

First, we can make rules for if a Concept/Detail is "found" or "not found."

Concept/Detail Configuration: We can make the AI take a few actions with Concepts. What's great is that as we click these different actions, we can see them updated in the Preview Tab of the far-right column.

• Alert: These will be in red, and the AI will warn you that something does not match your playbook. For example, liability may not be met, or the notice period may be shorter or longer than we want.
• Pre-Check: These will be in blue, telling the AI to let you know if something matches your playbook. Such as Governing law being correct, liability matching a certain number
• Escalate: This relates to our CLM integration and allows for an escalation to another team member if the AI finds something. For example, an escalation to a specific finance member can occur if liability exceeds a certain amount.
• Add: This will cause the AI to insert text into the document if something is "found" or "not found."
• Replace: You can have the AI replace specific text if it occurs. For example, if the governing law is Texas, we can have the AI replace it with "Delaware" or "State of Delaware."
• Delete: The AI can delete specific text if it is found. For example, we may not like Entire Agreements Clauses in our contracts, and we can have the AI remove them.
• Comment: Sometimes, we want the AI to leave a message for the reviewer; in this case, we can have the AI leave a comment on a subject that matches our playbook.

In-App comment: Leave quick notes and guidance that are viewable on the Issue without diving into the Details page.