Devs are spending only 16% of their time coding. Atlassian is engineering AI to reclaim the rest

The rise of AI coding agents has commoditized code generation, exposing a deeper challenge for software teams: the non-coding friction that consumes the vast majority of a developer’s day. Fixing that gap — not writing more code faster — is fast becoming the defining developer experience opportunity of this AI cycle.

Atlassian Corp. is betting its AI-native software development lifecycle strategy on exactly that premise. The enterprise collaboration software company’s own findings suggest that developers spend only a fraction of their time actually writing code — with the remaining lost to tasks such as finding information and context-switching between tools, according to Ming Wu (pictured), head of engineering for developer AI at Atlassian. That imbalance is the core problem Atlassian’s developer AI team is engineering against.

“Developers actually only spend 16% of time actually doing coding,” Wu said. “Our mission is improving developer productivity. The key for improving that is actually solving … the things that slow them down, like doing code review, digging around for information, being put into alignment and redoing the work and trying to clarify the requirements.”

Wu spoke with theCUBE’s Christophe Bertrand and Alison Kosik at the Atlassian Team event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how AI is reshaping the developer experience, where friction persists across the software development lifecycle and what Atlassian is building to address it. (* Disclosure below.)

Developer experience in the AI-native era

The industry’s initial wave of AI investment overwhelmingly targeted code generation, but that focus is rapidly expanding. Atlassian’s Rovo Dev platform, which connects Jira, Confluence and Bitbucket into a shared context layer, reflects a broader push to apply AI across the full software development lifecycle rather than just coding in the moment. The hardest part of that effort is not the tooling itself, but the context infrastructure underneath, Wu explained.

“The core part of those things — to deliver the quality solution with all those tools — is really the shared awareness and the context behind the scene,” Wu explained. “We call it context engineering. Without the right context, no matter how smart you are, even for humans, you wouldn’t be able to do the right thing.”

Beyond context, the changing developer role is reshaping how organizations think about team structure altogether. The traditional handoff between product managers and engineers — where requirements were written, handed over and then built — is being compressed, according to Wu.

“What’s happening now is that the line between traditional developers and other disciplines actually is somewhat blurred,” Wu said. “Everyone is getting to [be] more cross-disciplined and understand each other’s domain a bit more.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the Atlassian Team event:

(* Disclosure: TheCUBE is a paid media partner for the Atlassian Team event. Sponsors of theCUBE’s event coverage do not have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE