AI can sound confident about almost anything. The challenge is making sure it’s right. At Atlassian, Content Technologists are building the systems, standards, and structures that help AI deliver better outputs, removing the burden from our customers of knowing what information to trust.

In a world of AI noise, the most valuable currency is accurate, targeted content that gives customers full confidence they have the correct information, not slop. At Atlassian, we have always held a super-high bar on the information we publish. Content design and quality are part of our DNA. From our early days as technical writers, to the rise of content design and strategy, we’re proud of our seat at the table designing branded, accessible, and clear information across every Atlassian app surface.

As Content Designers, we have always cared deeply about the way our customers navigate, find and act on information across our apps. With AI now firmly in the centre of that experience, the role has gone deeper into the technology stack than ever before. Our craft is now embedded in the system prompts and governing knowledge bases that keep AI honest. And executing on that takes a special kind of role: someone equally comfortable in a CMS or a CLI. Enter, the Content Technologist.

Introducing the Content Technology role

These new skills require new role descriptions, which help the business and our craft partners understand where we sit in the new picture. As Atlassian software becomes more intelligent, we’ve met the moment by introducing new specializations that better position our role ‘under the hood’ where the real language action is being shaped. Content Technology is one of these new roles. And it’s a great example of how our impact is moving from designing the surfaces, to designing the systems that unlocks teams at scale in an AI era.

Haven’t we been here before?

Working on content infrastructure beneath the surface isn’t new; we’ve called it Content Engineering, Content Architecture, Content Taxonomy, and similar names. But AI raises the role’s importance. Large Language Models can produce fluent content, yet fluency doesn’t guarantee accuracy, consistency, or usefulness. An AI system is only as reliable as the content, context, tools, and constraints that surround it. Without these, users must bear the cognitive load of parsing information and working out what’s useful, and what’s not. This is where the Content Technologist comes in.

What Content Technologists do

Content Technologists help make AI-native experiences more grounded, observable, and governable. They turn content expertise and human judgement into structural guidance, constraints, and signals that machines can reliably act on. Examples include designing the guardrails and fallbacks that help an experience recover safely; adding metadata that helps an agent distinguish one concept from another; or making sure that a single piece of content can be assembled dynamically for a particular customer or journey — or be easily retrieved by an AI agent.

The role is not a replacement for engineering, rather its complementary and, as always, close partnership is key. The Content Technologist ensures that a technically elegant system can support meaningful, usable content. Engineers ensure that the system is realistic, reliable, and deployable. Together, these perspectives produce better solutions than either discipline can create alone.

Examples of impact

20% less token spend and twice as fast to retrieve

With Atlassian Design embracing AI-native ways of working, we needed a way to teach Atlassian’s AI Rovo and third-party AI tools to sound like Atlassian. Our Content Technologists built a shared repository containing 140 verified, authoritative content standards and made them available via Rovo Agents, MCP and Skills. This project not only empowered design teams to create content with AI, it also improved technical performance resulting in lower token spend and faster response times.

The impact in numbers:

  • ~21% lower cost
  • ~2× faster
  • ~20% fewer tokens
  • ~60% fewer tool calls

Syncing developer docs with code

A duo of Content Technologists designed a repeatable delivery framework for customer-facing developer documentation published on https://developer.atlassian.com/. Triggered by a request in Slack, our tailored Rovo agent now autonomously reviews the ask, creates a first draft, raises a PR, and tags a human for review – all staying within the systems where engineers already work! This has been a huge win for meaningfully reducing time to first draft, and keeping critical developer docs in sync with code.

Improving AI accuracy by 30%

Our Content Technologist working on support.atlassian.com ran a controlled experiment to test whether restructuring self-help documentation into a 4-layer content model with taxonomy metadata, produced measurably better AI outcomes than ‘flat’ content. By restructuring our support content into discrete, taxonomy-enriched knowledge blocks, AI answers grounded in these docs were 30% more accurate and ~45% more consistent than the same AI on today’s flat docs, for ~9% more cost per query.

The content design role revolution

With Content Technologists focusing on the substrate (the models, schemas, pipelines, platforms and behaviors that determine what content systems and agents know and how they use it), we’ve introduced other specializations that recognize our changing role in a ‘headless world’.

Model design is another new role that is shaping existing models, and training new ones, so that customers can more easily build trust in our AI experiences. We’re also using our skills to train open models to offload high-frequency, bounded tasks (like summarize page) so that they don’t require expensive frontier intelligence to execute. It’s exciting stuff and we’re learning a lot every day!

At its heart, this work is still driven by our love of language, semantics and our commitment to quality information. The craft we’ve built through years of making content clear, consistent, accessible, and useful is now helping us shape the systems behind AI: the structures that ground it, the standards that constrain it, and the feedback loops that improve it. By putting those skills to work, we can help ensure AI behaves the way our customers need it to: with accuracy, context, reliability, and care.