GenAI is everywhere these days, but Amazon Web Services is considered a laggard by some. In fact, it's still early days and the market is still messy. On Wednesday, AWS may have just given generative AI more credibility with a new tool called App Studio.
App Studio promises to help you create enterprise software applications from written instructions, and that's right: just describe the programs you need, and AWS will write the code for you, without the need for professional developers.
“App Studio is for techies who have technical expertise but aren't professional developers, and enables them to build enterprise-grade apps,” Sriram Devanathan, GM of Amazon Q Apps and AWS App Studio, told TechCrunch.
Amazon defines an enterprise app as an app with multiple UI pages that can pull data from multiple data sources, perform complex operations like joins and filters, and embed business logic.
The course is aimed at IT professionals, data engineers, enterprise architects, and even product managers who may not have coding skills but have the corporate knowledge needed to understand what software applications their company will need. The company wants these employees to be able to build applications by describing the applications they need and the data sources they want to use.
Examples of application types include an inventory tracking system or an invoice approval process. Users start by entering a name for the application, calling a data source, and describing the application they're building. The system provides some sample prompts, but users can also enter ad-hoc descriptions if desired.
It then creates a list of requirements for the application and what it will do, based on the description. The user can refine those requirements by interacting with the generative AI. In that respect, it's similar to many no-code tools that came before, but Devanathan says it's different.
“What makes App Studio different is how it uses Gen AI to significantly shorten the learning curve. I would say there's almost no learning curve. With typical no-code tools, you have to understand the paradigm. You have to understand the visual interface. You have to develop some expertise over time,” he said.
Once an application is complete, it goes through a mini DevOps pipeline and undergoes testing before going into production. When it comes to identity, security, governance, and other requirements for the application that a company deploys, administrators can link to existing systems when setting up App Studio. Once the application is deployed, AWS handles everything on the backend on behalf of the customer, based on the information entered by the administrator.
Devanathan says App Studio uses multiple models, including the Amazon Titan and Anthropic, depending on the job.
App Studio will be available in preview starting Wednesday.