Atlassian Intelligence is a powerful tool designed to supercharge your team's efficiency and accelerate work in Jira. With AI's generative capabilities and the wealth of teamwork data from Atlassian’s ecosystem, the possibilities are virtually endless. But, as with any advanced tool, the question arises - how can you tap into its full potential?
That's where ROC comes into play, and no, I'm not talking about rocking out with an electric guitar 🎸 (though that would be fun). ROC is an acronym that stands for Role, Output, and Context. It's a handy framework that can guide you in making the most of AI in your Jira issues.
Think of ROC as the key to unlocking the capabilities of Atlassian Intelligence. 'R' stands for the Role you want AI to take, 'O' stands for the Output you want AI to provide, and 'C' stands for the Context within which AI should operate. By clearly defining these three components, you can guide AI to better understand and respond to your prompts.
Click on the Atlassian Intelligence icon
AI will give you a few prompts to help you start.
AI prompt engineering is the process of crafting, refining, and optimising input prompts to guide generative AI systems in producing specific, high-quality outputs. When working with generative AI models, prompt engineering plays a crucial role in helping the system understand the nuances and intent behind your queries.
ROC 🎸 your AI tasks with adding [Role] + [Output] + [Context]
Example 1: As [Role] + [Task]
Prompt A: [Task] | Prompt B: As [Role] + [Task] |
How many story points do you think this will be and why? |
You are an engineering manager, how many story points do you think this will be and why? |
AI to act as a second opinion to help with effort estimates. In prompt A it does not return a result but in prompt B it provides an estimate.
Example 2: As [Role] + [Task] + [Output]
Prompt A: [Role] + [Task] + [Output] | Prompt B: [Role] + [Task] + [Output] |
You are a test manager, consider these points to write test cases |
You are a test manager, consider these points to write test cases formatted as a table |
AI to provide response in the output format you need.
Example 3: As [Role] + [Task] + [Context]
Prompt A: [Role] + [Task] + [Context] | Prompt B: [Role] + [Task] + [Context] |
You are a product manager, suggest detailed list of edge cases |
You are a product manager, suggest detailed list of edge cases and consider this will be for an app on iOS or Android |
Provide the context the experience will be an app.
Want more? Checkout the docs for more, like natural language search to JQL.
Don't have access to Atlassian Intelligence in Jira? Read these docs how to enable AI.
ROC 🎸 on with AI! Share in the comments, how do you prompt AI in Jira?
Thanks a lot for the useful prompts. I am totally struggling with this one.