Base44 this week unfurled a cloud-based platform that leverages artificial intelligence (AI) to foster greater collaboration between software engineers, application designers and product management teams.

Yoav Orlev, chief product officer for Base44, said at the core of the Base Code platform is a harness the company developed that enables multiple AI models to be used to analyze a codebase hosted in a GitHub repository. Those insights are then used to spin up a sandbox environment for application developers and designers based on service requirements discovered by the AI harness, such as databases that need to be spun up, said Orlev.

Once configured, designers and application developers can interact with code in a workspace that is based on a familiar Figma-style interface that many of them already use to build web applications, he added.

The overall goal is to reduce the level of friction that teams of developers and designers often experience when iteratively building applications in a way that also makes it simpler for software engineers to review the pull requests (PRs) those teams generate, said Orlev.

As application development continues to evolve in the AI era, the line between application developers and designers will continue to blur, especially as AI coding tools enable more so-called citizen developers to participate in the workflows used to build and deploy applications, noted Orlev.

As such, DevOps teams will need to deploy a cloud-based platform to supervise those interactions without having to necessarily participate directly in each workflow, said Orlev. Otherwise, there will simply be too many bottlenecks slowing down the pace at which applications are built and deployed no matter how much code is generated using AI tools, he added.

Hopefully, there will come a day when the PRs being generated by application developers and designers will be reviewed more by AI agents. Right now, many DevOps teams are simply unable to review all the code being created, which ultimately means there is more code running in production environments that has not been properly vetted than anyone cares to admit.

It’s not clear at what pace organizations are revisiting how they build and deploy software in the AI era, but at this juncture it’s more a question of when rather than if. While more code than ever is being generated, the DevOps pipelines currently relied on to validate and deploy all that code are being overwhelmed. Ultimately, DevOps teams will need to embrace some type of agentic engineering approach to managing a software development lifecycle (SDLC) that will encompass individuals of varying levels of application development expertise.

The pace at which DevOps teams make that transition will naturally vary, but laggards will soon find themselves falling behind rivals that have embedded AI across their SDLC. Handing out AI coding tools to developers and designers is only the first step in re-engineering how software is not just built but also delivered, maintained and updated in an era where the number of applications being deployed will soon approach a level of scale that not too long ago would have been considered unimaginable.