The experimental phase is over: Atlassian bets on DX to deliver AI ROI

All too often when it comes to AI, enterprises get stuck in a sort of experimentation limbo as deploying AI tools that deliver tangible ROI continues to be a struggle.

Atlassian aims to get them over that hurdle with its definitive agreement to acquire DX, an analytics platform that helps enterprises track their engineering teams’ performance and identify bottlenecks in development and deployment.

The acquisition will allow Atlassian to offer an all-in-one platform with integrated collaboration, project management, and developer productivity measurement tools to help enterprises move from AI testing to deployment.

“What a lot of leaders are saying is, the time for experimenting has passed,” Thomas Randall, research director at Info-Tech Research Group, told Computerworld. “It’s time to now actually put together the solid business case, and greater observability helps.”

Augmenting Atlassian’s ‘System of Work’ strategy

The DX platform was developed to help engineering leaders measure, benchmark, and improve developer productivity. Its tools will fit into Atlassian’s “System of Work” platform, which is designed to speed up processes by connecting business, service, product, leadership, and software teams across an enterprise.

According to the company, DX’s integration into Atlassian’s software development lifecycle (SDLC) will help customers measure AI adoption and impact, pinpoint and address bottlenecks, and gain visibility into developer experience, productivity and system health.

Atlassian has been investing significantly in the software space, with new offerings in Bitbucket Pipelines and Rovo Dev augmenting its Jira, Bitbucket, and Compass tools. The company is also staking a claim in the nascent AI-browser space with an acquisition of The Browser Company of New York.

“For any technology-driven organization, software teams are pivotal for your continued success,” Mike Cannon-Brookes, Atlassian CEO and co-founder, said in a YouTube video accompanying the DX acquisition announcement. “These teams are at a critical moment. AI is transforming how software teams work. This creates both incredible potential, and a whole new set of questions.”

Atlassian has 300,000-plus global customers, and a significant portion of DX’s customers, including Dell, BNY, Chime, Pinterest, GoodRX, Etsy, Vanguard, Pfizer, and Docker, already use Atlassian products.

“By joining forces, Atlassian and DX will help hundreds of thousands of software teams amplify their impact,” said Cannon-Brookes.

Providing insight into friction points, developer satisfaction

Organizations still struggle to pinpoint use cases for AI in their internal workflows and products, Info-Tech’s Randall noted.

“It’s been a bit more of throwing a lot of things at various pieces, or bottom up approaches from departments trying out different things,” he said. “A lot of time and money has been lost because too many projects and tests are going on.”

The integration of DX into Atlassian’s platform will offer observability and context, so teams aren’t just getting caught up in endless testing and experimenting loops, he said. “The core benefit is far greater visibility into where their work is slowing down. They can figure out which investments are actually working, which are having weaker impact.”

For instance, they can gain insight into code reviews or test delays. Many software teams are also adopting code assistants and automated testing tools, which DX can benchmark, Randall noted. They can also use it to gauge developer satisfaction and friction points, allowing them to set norms around measuring progress, and identify best practices and new opportunities for automation.

“It will help reduce technical debt by improving flow, lowering cycle times, reducing delays, expediting incident resolutions and risk management,” said Randall.

A ‘nice perk’ for existing Atlassian customers

The integration is a “nice perk” and “value add” for Atlassian customers, Randall contended. The company can now offer a “comprehensive package” featuring build and run software, matrix and feedback loops.

Enterprises must always be mindful about execution, he emphasized. How well will this integrate into their change management, such as by convincing engineering teams to adopt certain metrics? These tools surveil behavior, so it’s important to consider developer personas and identify what will motivate them and help them do their best work, he pointed out.

With any AI project, enterprise leaders should consider their core business goals, and objectives and key results (OKRs), then determine how AI tools can support them, rather than just adopting AI for the sake of it, said Randall.

“Because,” he noted, “a lot of the time it’s, ‘do you even need AI for that? This could just be straightforward automation with a rules-based system.’”