In this article, we will walk through the best practices for creating matters using the AI Agent, especially within the context of a Contract Review use case. This will include guidance on structuring your Action Name, Action Description, and Matter Layout fields for optimal clarity and efficiency. We will also explore some common pitfalls to avoid during matter creation.
AI Matter Creation Components
When creating matters in the AI system, it's important to structure each component in a way that ensures clarity, consistency, and efficiency. Let's break down the key components of matter creation:
1. Action Name
The Action Name is the first element of your matter creation process, and it should clearly define the task at hand. The name should follow a consistent format to help users quickly understand what the action is for.
Best Practice: Use a format like:
Create [MATTER TYPE]
Examples:
-
Create contract review
(Used when initiating a contract review process) -
Create IP Request
(Used for intellectual property inquiries or requests) -
Create Employment & Labor Issue
(Used when addressing employee-related matters, such as workplace issues)
2. Action Description
The Action Description provides context for when the matter should be created. It should start with the phrase "Call this tool when..." to indicate the circumstances under which this action should be triggered.
Best Practice: Ensure the description clearly outlines the types of requests or situations that would require matter creation.
Examples:
- "Call this tool when the user makes a request related to any of the following: employee contracts, disputes, benefits, and compliance with labor laws." (For a Contract Review matter type, this description ensures that the system understands when to initiate a contract review)
- "Call this tool when the user asks for a contract or when the user makes a request that relates to a contract." (Another example for Contract Review, ensuring it's triggered whenever anything related to contracts is mentioned)
- "Call this tool when the user mentions they have accepted a gift, need approval for a gift, or want gift(s) reviewed." (For a Gift Approval use case, showing a different matter type)
3. Matter Layout Fields
Matter Layout fields are important for capturing specific details about the matter. These fields help provide clarity and ensure the right information is recorded for each matter created.
Best Practice: Use clear and concise field names and descriptions. In most cases, descriptions are not necessary unless the field name requires further explanation for company-specific or unique fields.
Example for Contract Review Matter:
- Field name: Active Business Decision Involvement
- Field description: Indicates whether the gift provider is currently involved in any business decisions, negotiations, or contracts with the company that could create a potential conflict of interest.
4. Example of Matter Creation for Contract Review
In the Contract Review use case, let’s illustrate the complete matter creation process using the guidelines provided above.
Example 1: Action Name
- Action Name: Create contract review
Example 2: Action Description
- Action Description: "Call this tool when the user asks for a contract review or when the user mentions anything related to a contract, such as reviewing terms or making changes to an agreement."
Example 3: Matter Layout Fields
-
Field name: Contract Type
Field description: Type of contract being reviewed, such as employment, supplier, or partnership agreement.
-
Field name: Review Priority
Field description: The priority level for the contract review (e.g., high, medium, low).
-
Field name: Review Deadline
Field description: The date by which the contract must be reviewed or finalized.
Things to Avoid
When creating matters, especially for something as crucial as contract reviews, it's important to avoid potential pitfalls that could lead to inefficiencies or confusion in the system.
1. Too Many Actions
While the AI system supports up to 10 "Create Matter" actions, having too many actions can make the process cumbersome and less accurate. If you notice that accuracy is an issue, consider reducing the number of actions.
Recommendation: Limit your actions to only those that are necessary for specific, distinct use cases.
⚠️ Examples of Overload: Multiple Actions for Contract Reviews
Let’s say you create separate actions for different types of contract reviews:
- Create Sales Contract Review
- Create Vendor Contract Review
- Create NDA Review
- Create General Contract Review
At first, this may seem like a good way to categorize contract types. However, these actions all serve a similar purpose—reviewing a contract—just with slightly different contexts.
This can lead to the following issues:
- Overlapping logic: The AI may struggle to differentiate between when to use “Sales Contract Review” versus “General Contract Review,” especially if the user doesn't specify the contract type.
- Lower accuracy: Too many similar options can reduce the accuracy of AI matter routing, resulting in the wrong action being triggered.
- Harder to maintain: Any changes in the contract review workflow would need to be applied across all similar actions.
✅ Recommended Approach
Instead, consider consolidating into one action:
- Action Name:
Create Contract Review - Action Description: "Call this tool when the user requests a contract to be reviewed, including but not limited to sales agreements, vendor contracts, NDAs, or other legal documents."
This approach simplifies the logic, improves AI accuracy, and reduces maintenance effort—while still supporting multiple contract types.
By following the AI Matter Creation Best Practices, you can ensure a smoother and more efficient experience for creating matters, particularly in complex use cases like Contract Review. Keep the action names clear and concise, provide detailed descriptions that define the context, and avoid unnecessary overlaps or excessive actions. Properly structured matter fields also help in gathering essential information, ensuring your team has what they need to address any legal issues promptly and efficiently.
By adhering to these best practices, your team can leverage AI effectively to streamline contract review processes and enhance overall productivity.