This feature is in open beta release. It’s available for use but still in active development and may change.
With ZapAI, you can use artificial intelligence to analyze and parse your text.
You can use ZapAI to do things like:
- Parse new email leads.
- Route incoming messages to the right team.
- Analyze customer reviews for positive or negative sentiment.
- Identify spam content.
- Extract specialized information from freeform content.
Parse lead from freeform text
Use the Parse Lead from Freeform Text action to extract lead data for use in later Zap steps.
- In the Zap editor, click the Action step, or click the plus + icon to add an action to your Zap.
- Select ZapAI.
- Click the Event dropdown menu and select Parse Lead from Freeform Text.
- Click Continue.
- In the Action tab, click the Freeform Text Input field and select a field from the previous step you want to parse. This text field can extract the following:
- First name
- Middle name
- Last name
- Phone number
- Address line
- City
- State
- Zip code
- Click Continue.
- In the Test tab, click Test step. If the action is successful, the lead data will be extracted into separate fields. If no value is present for a field, ZapAI will return a null value.
Route messages
Use the Smart Message Router action to determine where to send your messages. This action returns the name of the route for use in later Zap steps.
- In the Zap editor, click the Action step, or click the plus + icon to add an action to your Zap.
- Select ZapAI.
- Click the Event dropdown menu and select Smart Message Router.
- Click Continue.
- In the Action tab, click the Message field and select a field from the previous step you want to route.
- In the Routes section:
- In the first column, specify the route's name, such as a person, group, or team.
- In the second column, you can add multiple keywords or phrases, separated by commas, to describe each route.
- Click Continue.
- In the Test tab, click Test step. If the action is successful, the correct route will be returned. If your message doesn’t exactly match the route descriptions, ZapAI will determine which route is most appropriate.
In this example, a Zap was set up with the following routes:
- Support
- EPD
- Marketing
- Accounting
- Research
The Zap received this input data: “Need some numbers around the increase in sales”. From this, ZapAI determined that “Accounting” was the appropriate route.
Parse any freeform text
Use the Parse Any Data from Freeform Text action to parse any text input. Enter custom examples to train the AI to return the data you need to use in later Zap steps.
- In the Zap editor, click the Action step, or click the plus + icon to add an action to your Zap.
- Select ZapAI.
- Click the Event dropdown menu and select Parse Any Data from Freeform Text.
- Click Continue.
- In the Action tab, click the Freeform Text Input field and select a field from a previous step to parse.
- You can add up to 5 examples per action. To add an example:
- In the (Example 1) Freeform Text field, enter example text that might be received from the Freeform Text Input field.
- In the first column of the (Example 1) Expected Data field, add a word or phrase that describes what you want the AI to parse.
- In the second column of the (Example 1) Expected Data field, add a word or phrase that the AI should output.
- Click Continue.
- In the Test tab, click Test step. If the action is successful, the correct route will be returned.
To produce the correct response, you may need to add additional examples to train the AI to produce the correct response. Re-test each time an example is added to test the response.
In this example, ZapAI will analyze the input text for:
- “Customer sentiment”, like whether it’s a positive or negative sentiment.
- “Needs reply”, whether a reply is needed to this customer.
ZapAI limits
- You can enter approximately 10,000 characters for all data inputs and outputs combined.
- ZapAI doesn’t store data from previous actions to learn from. Each time a ZapAI action runs, the AI is reset to a blank state.