AI Builders Week
The core theme running throughout the entire Atlassian AI Builders Week guide is to explain “how product managers (PMs) and designers become not simply ‘developers’ who directly write code, but leaders (Steering) who use AI to maximize work efficiency and guide the entire team in the right direction.”
1. "Do not row; steer the ship"
- Expansion of roles: In the AI era, the roles of PMs, designers, and engineers do not merge into one (`AI Builder`), but rather the circle of each discipline expands further.
- Resolving bottlenecks: Thanks to AI, engineering productivity has increased explosively. At this time, if PMs and designers are only doing direct coding, bottlenecks arise on the planning/design side.
- Setting direction: The most important role of PMs and designers is not directly writing code, but setting direction and increasing decision-making speed so the team can understand and execute the right product direction and context.
2. Practical guide to AI use and on-the-job work for PMs and designers
💡 High-value use for PMs
- Multi-step work automation: Going beyond simple chatbot questions, guide systems to handle end-to-end workflows by linking files, tools, and code.
- Examples of practical application:
- Compile Jira, Confluence, and Loom data to write a weekly work summary
- Establish product strategy and roadmap through conversational prompts (using the RICE framework)
- Automatically create Jira Product Discovery (JPD) tickets and manage the backlog
- Create a working prototype based on backlog ideas
🎨 Use and capability enhancement for designers
- Democratization of design decisions: Excellent designers in the AI era do not stop at quickly generating designs with AI; they build guidelines and systems so the entire team can make good design decisions.
- Strategy and product framing: Beyond the existing design domain, expand capabilities so that you can broadly participate in areas such as strategy formulation and writing product frameworks, up to the PM domain.
3. Practical tips for AI agents and token optimization
- Using Rovo CLI: Educate teams to actively use a terminal-based autonomous agent that is not merely an AI for simple answers, but has the ability to read local computer files, integrate with Atlassian tools, and retain permanent memory.
- Token budget management: Tokens should be approached not as an unlimited resource but as a 'budget'. Guide so that the context window is maintained at around 50~70% to prevent hallucinations or performance degradation in the agent.
- Skill best practices: Recommend a culture of structuring repeated prompts (`skill.md`) and sharing/reusing them through shared repositories with the team and the entire organization.
