Airtable, a no-code platform for building workflows and apps, today announced that it’s launching a beta program and expanding access to Airtable AI, allowing enterprises to develop and deploy generative artificial intelligence applications for business workflows and productivity.

Using Airtable AI, companies will be able to integrate generative AI across their organizations and workflows without the need for advanced technical knowledge. Generative AI is a type of AI that can produce new text based on user prompts, which can be used to create marketing content, summarize customer feedback and automate communication.

Airtable AI was initially introduced in a limited alpha during May for a small number of participants. The beta program will start with participation from a larger number of retail, services, technology and media companies, then will scale up to the thousands of organizations currently using the platform later this summer. It will allow them to use the AI building block to experiment with AI capabilities using their own data and business workflows.

“The current generation of large language models are already highly capable of advanced reasoning and creativity tasks and can be applied to a broad range of knowledge work,” said Airtable co-founder and Chief Executive Howie Liu. “Enterprises can now use Airtable’s no-code approach to immediately integrate these powerful models in the context of their existing data and workflows, making it possible to rapidly build and deploy highly engaging, AI-powered apps.”

Using Airtable’s platform, knowledge workers and employees can simply pick and choose the data they want to feed it and then have the generative AI do the rest of the work Anthony Maggio, senior director of product management, told SiliconANGLE. That makes it simpler to use and works directly with the company’s data through Airtable’s database and automation platform.

Airtable AI is launching initially with access to OpenAI LP’s foundational models, including GPT-3.5 and GPT-4, Maggio said. However, the AI building block itself is designed to be model-agnostic and in the future will be able to access enterprise-specific models.

“It really gives teams the ability to harness the power of these new AI models within their own workflows,” said Maggio. “So the data that they’re storing or they’re capturing as part of their workflows can be pushed into this new AI building block and actually perform steps of their workflow or accelerate pieces of their process.”

Too often, AI integrations aren’t especially intuitive or require technical expertise to make them work within business workflows and that’s not ideal, Maggio added, or they’re so broad or generic that they’re not particularly useful for teams with specific needs. “Because it’s integrated as a building block into our platform as a whole, we really see this as the best way for enterprises to be able to give their employees the ability to experiment with utilizing this new technology,” he said.

Examples of how users put the new AI building block to use during the alpha phase included sentiment analysis of customer feedback – such as positive or negative – and other classification systems. Others put the Airtable AI system to use in reducing tedium for workers for content generation and categorization.

The service can be used to transform product development by automatically tagging and categorizing large amounts of user research by triaging commonly asked questions to generate potential Frequently Asked Questions documents, for products or services based on customer feedback. The same thing could be used to fine-tune resources and product descriptions based on customer responses.

Although right now Airtable AI only supports text generative AI, Maggio said both image generation and image classification and tagging and search are on the radar of the company as potential future investment areas.

Customers interested in joining the waitlist for the expanded beta can sign up on the company’s website portal starting today. Airtable AI is planned to become an optional planned feature available for customers later this year with details on pricing and availability to be announced soon.

Image: Airtable