Customize your AI Agent's generative content with Guidance

You can customize your AI Agent's behavior by adding custom instructions, which are called pieces of Guidance, to your AI Agent. Then, when your AI Agent is sending messages to customers, it takes your Guidance into account.

Understand how Guidance works

Guidance provides you with a lot of freedom to change how your AI Agent speaks to your customers! There are some hard limitations about the Guidance you add to your AI Agent:

• You can have up to 40 active pieces of Guidance in your AI Agent. In addition to these active pieces of Guidance, you can also have inactive drafts that you're working on.

• Each piece of Guidance can be up to 300 characters long.

• Your Guidance must be in English.

Additionally, there are some things you should keep in mind to ensure that the guidance you add to your AI Agent performs well:

• Make sure your pieces of Guidance don't conflict with each other.

  For example, you could have two pieces of Guidance to tell customers about both a spring and summer sale. If both of them are live at the same time, you can't guarantee that the correct piece of Guidance will apply when a customer is talking with your AI Agent.

  Ada takes away the guess-work by letting you know if a newly added piece of Guidance conflicts with any existing piece of Guidance.

Beyond these limitations, you can be as creative as you want!

Add a piece of Guidance to your AI Agent

Here's how you can add a piece of Guidance to your AI Agent. As soon as you publish the piece of Guidance, your AI Agent can start considering it when generating responses to your customers.

If required, you can restrict pieces of Guidances to certain users, based on information your AI Agent collects about your users and saves in variables.

note

You can only use variables your AI Agent can collect through your browser, or that you collect in a block and allow to be available outside of the structured content the block is in. You can't use variables your AI Agent collects using Actions.

1. On the Ada dashboard, go to Training > Guidance, then click New Guidance. The Add Guidance page opens.

2. Under Title, add a title for your new piece of Guidance.

   note

   This title is just for your reference; your AI Agent never sends it to the LLM.

3. Under Instruction, add the new rule you want your AI Agent to follow.

4. Under Set availability, choose which customers you want the piece of Guidance to apply to.

   • To make the piece of Guidance available to all users, select Everyone.

   • To restrict the piece of Guidance to certain users, select Based on the following rules. A section expands where you can enter the logic your AI Agent will use to decide whether to serve the pieces of Guidance.

     1. Beside Where, in the Choose a variable list, select a variable.

     2. In the next dropdown, select an operator so you can define a relationship between the variable and the value you want to target.

        Understand comparison operators

        Comparison operators are logic statements that tell your AI Agent to match customer information that’s captured in the variable you’re using. The available operators vary based on the variable type you’re using:

        Operator	Variable types	Description
        Begins With	All text variables (including phone and email)	Match information in the variable that begins with certain text (partial match).
        Ends With	All text variables (including phone and email)	Match information in the variable that ends with certain text (partial match).
        Contains	All text variables (including phone and email) List variables	Match information in the variable that contains certain text in any position (partial match).
        Is	All text variables (including phone and email) Number variables	Match information in the variable that equals specific text exactly (exact match).
        Is Not	All text variables (including phone and email) Number variables	Match information in the variable that does not equal specific text exactly (exact match).
        Is Not Set	All text variables (including phone and email) Number variables List variables	Match if there is no information contained in the variable.
        Is Set	All text variables (including phone and email) Number variables List variables	Match if there is any information contained in the variable.
        Greater Than	Number variables	Match if the information in the variable is greater than a specific value.
        Less Than	Number variables List variables	Match if the information in the variable is less than a specific value.
        Is True	Yes/No variables	Match if the information in a variable is Yes (or True).
        Is False	Yes/No variables	Match if the information in a variable is No (or False).
        Does Not Contain	List variables	Match if none of the selected items contains this value.
        Is Equal To	List variables	Match if the customer selected a particular number of options in a List Option block.
        Is Greater Than	List variables	Match if the customer selected more than a particular number of options in a List Option block.
        Is Less Than	List variables	Match if the customer selected fewer than a particular number of options in a List Option block.

     3. In the Value field, enter or select a value for the variable that you want to use to target users.

     4. If required, add additional conditions.

        • By default, if you click the Add icon , new conditions will be added with "or" statements, meaning customers will see the pieces of Guidance if they fulfil any of your criteria.
        • You can click And to add additional conditions that must all be satisfied for your AI Agent to serve the pieces of Guidance.

5. To save your new piece of Guidance as a draft, click Save. If you save it as a draft, it won't affect your AI Agent content until you publish it.

6. To test your new piece of Guidance, click Test. A chat window opens, where you can type prompts and test whether the piece of Guidance returns the kinds of responses you want, even if it's still a draft.

   If you haven't saved your piece of Guidance yet, testing it saves it as a draft.

   note

   If your new piece of Guidance conflicts with any existing piece of Guidance, your AI Agent notifies you when you click Test. Conflicts are checked against both active and inactive pieces of Guidance.

7. To publish your piece of Guidance, beside the Inactive toggle at the top of the page, click the toggle so it says Active. As soon as you do, your AI Agent starts using your new guidance in the responses it writes to customers.

See the Guidance that factored into a conversation

On the Ada dashboard, when you go to Conversations, you can see transcripts of conversations that customers have had with your AI Agent. In the Details pane, under Guidance considered, you can see a list of Guidance that was active at the time of any given conversation, and use that information to fine-tune your Guidance. For more information, see Improve your AI Agent's responses by viewing past conversations.

Edit or delete a piece of Guidance

After you or a colleague create a piece of Guidance, you can go back to it so you can edit or delete it.

1. On the Ada dashboard, go to Training > Guidance.

2. Click the piece of Guidance you want to manage. The Edit Guidance page opens.

3. Make your required changes to the piece of Guidance.

   • You can edit the existing piece of Guidance and then click Save to update it and keep it active.

   • You can click Convert to draft to stop an active piece of Guidance from affecting customer conversations, without deleting the piece of Guidance from your AI Agent.

   • You can make changes to the piece of Guidance and then click Test to try it out in a conversation. Testing a piece of Guidance automatically saves it, but doesn't change whether it's a draft or active.

   • You can click the Delete button to permanently delete the piece of Guidance from your AI Agent.

Change availability settings for multiple pieces of Guidance

You can select multiple pieces of Guidance and quickly change whether they're available to all of your users, to only some of them, or delete your selected pieces of Guidance altogether.

1. On the Ada dashboard, go to Training > Guidance.

2. Select the pieces of Guidance you want to manage. Some options appear at the bottom of your screen:

   • To make the selected pieces of Guidance available to only some of your users, click Set availability. The Set Availability window opens.

     • To make the piece of Guidance available to all users, select Everyone.

     • To restrict the piece of Guidance to certain users, select Based on the following rules. A section expands where you can enter the logic your AI Agent will use to decide whether to serve the pieces of Guidance.

       1. Beside Where, in the Choose a variable list, select a variable.

       2. In the next dropdown, select an operator so you can define a relationship between the variable and the value you want to target.

          Understand comparison operators

          Comparison operators are logic statements that tell your AI Agent to match customer information that’s captured in the variable you’re using. The available operators vary based on the variable type you’re using:

          Operator	Variable types	Description
          Begins With	All text variables (including phone and email)	Match information in the variable that begins with certain text (partial match).
          Ends With	All text variables (including phone and email)	Match information in the variable that ends with certain text (partial match).
          Contains	All text variables (including phone and email) List variables	Match information in the variable that contains certain text in any position (partial match).
          Is	All text variables (including phone and email) Number variables	Match information in the variable that equals specific text exactly (exact match).
          Is Not	All text variables (including phone and email) Number variables	Match information in the variable that does not equal specific text exactly (exact match).
          Is Not Set	All text variables (including phone and email) Number variables List variables	Match if there is no information contained in the variable.
          Is Set	All text variables (including phone and email) Number variables List variables	Match if there is any information contained in the variable.
          Greater Than	Number variables	Match if the information in the variable is greater than a specific value.
          Less Than	Number variables List variables	Match if the information in the variable is less than a specific value.
          Is True	Yes/No variables	Match if the information in a variable is Yes (or True).
          Is False	Yes/No variables	Match if the information in a variable is No (or False).
          Does Not Contain	List variables	Match if none of the selected items contains this value.
          Is Equal To	List variables	Match if the customer selected a particular number of options in a List Option block.
          Is Greater Than	List variables	Match if the customer selected more than a particular number of options in a List Option block.
          Is Less Than	List variables	Match if the customer selected fewer than a particular number of options in a List Option block.

       3. In the Value field, enter or select a value for the variable that you want to use to target users.

       4. If required, add additional conditions.

          • By default, if you click the Add icon , new conditions will be added with "or" statements, meaning customers will see the pieces of Guidance if they fulfil any of your criteria.
          • You can click And to add additional conditions that must all be satisfied for your AI Agent to serve the pieces of Guidance.

   • To make the selected pieces of Guidance available to none of your users, click Set as inactive.

   • To make the selected pieces of Guidance available to all of the users configured in its availability settings, click Set as active.

   • To delete the selected pieces of Guidance, click Delete.

---

Have any questions? Contact your Ada team—or email us at [email protected].