Airtable automation actions: Generate with AI

Using AI in automations

With AI in Airtable Automations, you can combine the power of AI to build automated workflows such as:

• When negative customer feedback is submitted, summarize it, and share it in Slack with the appropriate team(s)

• Summarize team updates on Friday and send an email of the summary with the team every Monday morning

• Fill out an interface form, then have AI generate a guide in google docs and share the draft with content editors

Automations that involve the Generate with AI action will incorporate at least three components, but can certainly contain more depending upon the complexity of the workflow that you are wanting to automate:

1. An automation trigger that will tell Airtable to begin the automation and queue up the AI to run the prompt.

2. A "Generate with AI" action step with the prompt, randomness, and model sections configured.

3. One or more automation action steps that takes the content generated by the AI and does something with it - writes it to a record in Airtable, sends it in an email, posts it to Slack, etc.

Configuring an automation using the “Generate text” action

In this example, we will us AI in an automation that will send an email every Monday morning with a summary of all meetings last week.

Note

You'll want to adjust any sample settings listed in the walkthrough below to match the field, view, and table names in your base. Use this example, just as a reference to think about the overall structure of building this type of automated workflow.

To configure an automation using the “Generate text” action:

1. Open your Airtable home screen.

2. Open the base where you want to create an automation.

3. Click Automations.

4. Click + Create new.

5. Click + Create automation.

6. Click + Add trigger.

   1. Once clicked, the available triggers will display for you to choose from.

   2. Create a new automation with an "At a scheduled time" trigger set to run every 1 week on Mondays at 6:30 a.m. Test the trigger and move on once the trigger successfully tests.

7. Next, we will use a "Find records" action to find only meetings from the past week. In this example, we will set the "Table" to be Weekly meetings, then "Find records based on" a View, and then set the "View" to be This past week's meeting.

8. Next, you'll add the “Generate text” action step. Click Generate with AI and then Generate text.

9. Here you will configure the prompt, randomness, and model. Let's focus on the prompt since randomness and model are covered in this article. For the prompt, we want a mix of static text to give the AI model context along with the dynamic details about the meeting that the AI model can interpret to generate a summary.

10. You will also want to add a Send email automation action so that you can send the summarized meeting notes to your colleagues via email. Configure the "To" and "Subject" fields according to your needs. Then, in the "Message" you'll want to insert a token of the “Generate with AI” action's "Response." This response is the output that the AI generates based on the settings you configured in the prior steps of this automation.

11. Test and turn on your automation.

Configuring an automation using the “Generate structure data” action

The “Generate structured data” action is an extension of the “Generate text” action. This option allows you to define a specific data format for the AI to output, ensuring that the AI output is consistent and provides the exact fields you need. A few examples:

• You can use the data outputted by this action in later steps in the automation. For example, you can output lists and then loop over them, or output objects and use each individual property for different steps, such as creating or updating records.

• You can have the AI turn a raw text input (such as meeting notes or customer feedback) into specific fields to update a record with. Or, ask the AI to generate ideas or search the internet for information and create new records with the results.

A few tips for optimal usage:

• If you want the AI to output multiple items, you must include an “array” type in your schema. Arrays can be used as the input to a repeating group.

• If you have an array type in your schema, and the AI does not identify any items to populate it with, it will output an empty list

• You don’t need to instruct the AI to follow your schema in the prompt; it will automatically output data in the format you define, using the prompt for additional context as needed. You may not need any additional instructions in your prompt (only some source data) if the data clearly maps to your desired schema.

To configure an automation using the “Generate structured data” action:

1. Open your Airtable home screen.

2. Open the base where you want to create an automation.

3. Click Automations.

4. Click + Create new.

5. Click + Create automation.

6. Click + Add trigger.

7. Next you will add a structure data action. Click Generate with AI and then Generate structured data.

8. You will then configure :

   1. A prompt

   2. The output schema

   3. Toggle on/off the “Allow AI to access the Internet” option

   4. Randomness and model are covered in this article.

9. Add any additional automation actions. You can reference the output schema/structured data in these steps as needed.

10. Test and turn on your automation.

Limitations

• The total number of words in your query and response must be under a certain amount set by the AI model being used on the backend and Airtable’s automation system. This includes the words in the fields referenced by your query. If you exceed these limits in your query, you will see an error message. If the AI's response exceeds the limits, the text will be truncated.

  • Lower powered AI models: Up to ~12,000 words

  • Higher powered AI models: Up to ~90,000 words

  • In addition to the model word limitations, the prompt cannot exceed 64,000 characters.

• For the the “Generate structure data” action, there is a maximum of four levels of nesting for “Array” or “Object” outputs.

Note

For more information about model types and credit limitations is available in this article.

FAQs