Rovo Redefined

How I led the transformation of Atlassian’s Rovo from a basic chat feature into a central, workflow-embedded AI platform—driving a new design system, multi-team innovation, and a company-wide AI evolution.

Project Context

Context

When I joined Atlassian, Rovo—the company’s flagship AI product—was a constrained sidebar chat with an outdated visual language and a confusing system of hundreds of specialized agents. It wasn’t modern, it wasn’t competitive, and most importantly, it wasn’t deeply embedded into the user’s workflow.

Challenges

  • Outdated design system (ADS) that limited visual expression and interaction patterns

  • Fragmented agent model that forced users to know which agent did what

  • Chat restricted to a static sidebar with no inline or embedded intelligence

  • No cohesive AI vision across the Atlassian suite

My Role

I owned the product and design vision for Rovo Chat and the future of Atlassian’s AI experience. I:

  • Identified critical usability, design, and competitive gaps

  • Built the case for a new AI design system and a complete product strategy overhaul

  • Championed a company-wide realignment around AI

  • Ran a tiger team of Principal design leads from across the company to reimagine the end-to-end North Star Experience

  • Ran several tiger team workshops and strategic leadership reviews

  • Drove execution across 32+ teams to transform Rovo into a platform experience

Project Kickoff

Identifying the Problem

My first step was to run a deep competitive analysis of ChatGPT, Gemini, Anthropic, Replit, Cursor, Protopie, Vercel, and other emerging AI tools. I mapped:\

  • Interaction patterns

  • Chat modalities

  • Workflow integration strategies

  • Design system capabilities

  • Multimodal features

The gaps were clear—and severe. Rovo was falling behind.

Making the Case to Leadership

I synthesized findings into a strategic brief and presented it to our VP, highlighting:

  • The limitations of ADS

  • The fragmentation of our agent model

  • The missing cross-suite AI workflows

  • The risk of shipping a dated experience for GA

I proposed:

  • A full visual + interaction redesign

  • A new AI-specific design system

  • A cross-company Design Leadership workshop to realign all product teams

  • Dedicated design + product + engineering resourcing (tiger team)

The Workshop

Unifying 32 Teams Around a New AI Vision. To break the siloed and fragmented approach to AI at Atlassian, I planned, facilitated, and led a cross-company workshop with 32 teams across Jira, Confluence, Atlas, Compass, Ecosystem, and Platform.

Workshop Goals

  • 1

    Define a unified vision for Rovo

  • 2

    Redesign the entire chat experience

  • 3

    Create principles for the AI interactions across products

  • 4

    Decide where AI should live in user workflows

  • 5

    Validate the need for a new AI design system

The workshop

What I Delivered

  • Full workshop agenda & exercises

  • Vision-setting exercises

  • Co-creation sketches of new chat experiences

  • Workflow mapping for “Create with Rovo”

  • Alignment with design, engineering, product, and leadership

Workshop Outcome

The workshop produced:

  • A brand-new chat architecture

  • Agreement on three core chat modalities (inline, floating, full-page)

  • Concept for Create with Rovo

  • Foundational patterns for the Rovo Design System

  • A multi-quarter roadmap embraced across all teams

Launching the Product Work-streams

Based on the workshop outcomes, I kicked off and drove multiple major workstreams:

New chat modalities

A single sidebar chat was no longer enough.

I designed and led the definition of:

  • Inline Chat → AI appears where you work (in Jira tickets, Confluence pages)

  • Floating Chat → always-available, lightweight interaction

  • Full-Page Chat → deep work, research, drafting, exploration

Each modality had unique UX, component needs, and interaction patterns—which informed the new design system.

Create with Rovo Flow

The biggest leap forward: turning AI from chatting into creating.

This became:

  • A multimodal AI-driven drafting experience

  • A structured workflow for content generation

  • A cross-product creation flow for Jira, Confluence, and beyond

I defined:

  • Entry points across workflows

  • Drafting + iteration patterns

  • Content validation steps

  • Multi-agent orchestration

  • The publishing model

This work became the foundation of Atlassian’s new AI-powered workflow model.

AI Design Kit

I advocated for—and drove the formation of—the team that now owns Atlassian’s dedicated AI design system.

AI Design Kit now includes:

  • AI-specific UI components

  • Chat orchestration architecture

  • Prompt UX patterns

  • Inline intelligence patterns

  • Evaluation + feedback patterns

  • Multimodal patterns for image, tables, code

This became a new design organization within Atlassian Central AI Rovo.

Final Result

Rovo transformed into a unified AI platform

  • Modern, sleek, and competitive visual design

  • Chat embedded across multiple surfaces

  • Workflow-powered creation tools

  • Cohesive AI behavior and interaction patterns

  • A scalable design system for all future AI features

Rovo now feels:

  • Integrated

  • Intelligent

  • Elevated

  • Future-proof

  • And genuinely useful

Outcome & Impact

Company-Wide Impact

  • Rovo’s new experience became the centerpiece of the founders’ keynote at Team ’25 and Team ’25 EU

  • 35+ teams showcased demos built on the patterns I created

  • Sparked new hiring, new AI teams, and new design system org

  • Rovo became Atlassian’s flagship strategic priority

Business Impact

  • Significant revenue growth tied to AI add-ons

  • Investor excitement and shareholder momentum

  • Surge in customer engagement and product usage

  • Strong validation from analysts and the developer community

Organizational Impact

  • Shifted Atlassian from “AI as a feature” → AI as the product

  • Unified design + product around a cohesive AI vision

  • Established scalable infrastructure for all future AI tools