Amazon Web Services Inc. today introduced AWS App Studio, an artificial intelligence service that makes it possible to develop applications using natural language prompts.

The offering is designed to help companies more quickly build internal software tools for their employees. A marketing agency, for example, could use App Studio to create an application that helps staffers manage the creative assets used in ad campaigns. An information technology team might build a tool that tracks which employee workstations have to be repaired or replaced.

“Using natural language, any user with some technical experience can simply describe the application they want to build, and App Studio takes care of the development process, delivering an application that employees can start using immediately,” said Dilip Kumar, vice president of applications at AWS. 

The development workflow in App Studio starts with a text prompt. Workers must enter a brief natural language overview of what application they wish to build. That overview can be a one-sentence summary of the task for which the application will be used, or a more detailed description that outlines specific features. 

Based on the provided prompt, App Studio generates a detailed list of the features and interface sections that it plans to generate. The user can review the list and, if an important feature is missing, start from scratch by entering a new prompt.

App Studio takes a few minutes to generate a new application. After the initial version is ready, users can customize the program’s interface through a drag and drop interface. A built-in chatbot provides pointers on how to carry out changes.

Workers can optionally integrate their AI-generated applications with external services. A revenue tracking application, or example, could be configured to retrieve sales logs from a company’s Amazon S3 storage repository. App Studio provides integrations with hundreds of third-party services including popular cloud applications such as Salesforce. 

A built-in testing tool enables workers to check that their AI-generated software works as intended before deploying it. According to AWS, the testing tool makes it possible to generate sample data similar to the information the application will process in production. Users can review how the application interacts with the data to uncover technical issues.

After development is complete, App Studio makes each program accessible via a dedicated URL. Companies can apply role-based access controls to ensure that only authorized users can log into their applications.

AWS positions App Studio as a more economic alternative to existing low-code development tools. The service is available for free and customers are charged based on the amount of that time employees spend using published applications. According to AWS, this arrangement makes App Studio up to 80% more cost-efficient than rival services.

The service debuted at the cloud giant’s AWS Summit New York today alongside several other product updates. The Amazon.com Inc. added new features to Guardrails for Amazon Bedrock, a tool that helps companies prevent their AI applications from generating harmful output. AWS also made a customizable version of Claude 3 Haiku, a hardware-efficient language model from Anthropic PBC, available to customers on a managed basis.

Image: AWS