Airtable, already a repository of workflow data for its users, makes a good case study for AI as a productivity tool because it can bring together information from disparate cards and tables. It also brings Airtable closer to an all-in-one service, for everything from organizing teams to pushing specially tailored content to customers.

Airtable’s new AI-powered summarization and insights feature lets customers get a quick understanding of the information in their workspace and share insights with the rest of the team. It will also automatically apply categories and tags to information and route any action item to the relevant team. Within the same workspace, Airtable AI will also let users tap a button to generate emails or social posts. It will also translate data into other languages while keeping the same tone as the original content. The company did not specify which languages its AI features can understand.

Workplaces have been interested in introducing generative AI to help employees be more productive.

The company said the AI features are a “paid, opt-in feature that is only enabled if purchased and turned on.” Airtable will offer a free trial of 500 AI credits, enough to do 500 formulas or matching records, 10 blog posts, or 50 translations. After that, customers will have to pay an additional $6 per seat per month paid annually. Admins can also choose to activate the functionality for specific workspaces or all users. 

Airtable beta tested its AI features with over 1,200 organizations, including AWS. AWS also provided the AI models Airtable used to build its product through its model repository, Amazon Bedrock. Customers can access Airtable AI features through different AI models, using Airtable’s work with Amazon Bedrock. Currently, users can choose OpenAI’s GPT-3.5 and GPT-4 or Anthropic’s Claude model, but the company said more model options will be available.