What types of data does my organization contribute?

We’re updating how we use customer data so we can improve our apps and AI experiences for all customers. Learn more about this change

 

Your Atlassian organization contributes metadata and in-app data based on your data contribution settings.

Read this page to understand:

  • what metadata and in-app data include

  • how we safeguard your data contributions. More about how we protect your data

  • how your data contributions improve app and services for all customers

  • retention of contributed data

Metadata

Metadata is composed of content attributes and common patterns. Refer to this table to understand the composition of metadata.

Content attributes

Common patterns

Content attributes are statistical characteristics, numeric fields, and derivatives of your in-app data.

Content attributes used to improve apps and experiences for all customers:

  • don’t contain information that reveals health, location, workforce, or financial data

  • don’t reveal information that identifies an individual, because they are de-identified and aggregated before use

Some examples of content attributes include:

  • readability score or the complexity of Confluence page content

  • task classifications assigned to content, such as, “sales work item”

  • semantic similarity score, such as, how similar two Confluence pages are

  • numbers entered into Atlassian-created fields in apps, such as:

    • the story points assigned to a Jira work item

    • the end date of a sprint in Jira

    • the Service Level Agreement (SLA) of a Jira Service Management request

Common patterns are derived only from the following types of in-app data:

  • search queries and results

  • prompts and responses in Rovo Chat

  • configuration items you define (for example, custom fields in Jira)

To create common patterns that we use to improve apps and experiences for all customers, we extract data that is common across customers and omit data that is low in frequency and may be unique to your organization.

Some examples of common patterns include:

  • common words or phrases used in search queries across customers, such as, “vacation policy”, “product roadmap”, or “priorities”

  • common Rovo Chat prompt topics, such as, “What team is responsible for?” or “Recap team activity”

  • common configuration values, such as, a custom Jira work item type like “iOS bug” or “Laptop repair ticket”

  • common Rovo agent names, and description keywords, such as “Procurement helper”

Metadata helps us build and refine Atlassian app features based on how customers typically use them. With metadata, we can:

  • make tools like Jira and Confluence more intuitive and efficient

  • train our search model to improve the relevance of search results

  • provide accurate and context-rich root cause analysis recommendations so you can resolve incidents quicker and proactively prevent future issues

  • surface intelligent suggestions that accelerate tasks

In-app data

In-app data refers to content created by users within Atlassian apps. This can include things like:

  • titles and content in Confluence pages

  • titles, descriptions, and comments in Jira work items

  • custom emoji names

  • custom Jira or Confluence status names

  • custom workflow names

Before we use in-app data to improve apps and services for all customers, we de-identify and aggregate it. We remove information that directly identifies an individual, such as name or email.

In-app data helps us create better app experiences by learning how customers typically use Atlassian apps. These patterns can help improve search relevance or identify frequent tasks.

We use this data to develop features that:

  • provide more tailored recommendations and next steps in workflows. For example, instead of Rovo responding with “Do you need help with anything?”, Rovo might suggest “Shall I create a work item for this?”

  • return fewer dead ends when acronyms mean different things. For example, if you prompt “What does ML stand for?” in Rovo, you may receive a response that is relevant to your domain (Machine Learning in a software development context or Maximum Load in a manufacturing context).

Retention of contributed data

We may extract data that appears frequently across metadata or in-app data from all customers who contribute. This recurring data, similar to common patterns, will be de-identified and aggregated at a customer level, such that it is no longer associated with individual customers.

Data that is de-identified and aggregated at a customer level such that it is no longer associated with individual customers (including common patterns), may be retained for up to seven years.

Retaining this data, which has been de-identified and aggregated at a customer level and is common across customers, enables us to make more meaningful observations over longer periods of time. By unlocking deeper insights into customer behavior, we’re able to drive continual improvements to your overall experience, including by enabling Rovo to:

  • surface the most relevant results in response to prompts and queries

  • summarize relevant content more accurately and concisely

  • identify the best templates for creating new documents

  • learn which agentic workflows and follow-up questions lead to successful completions, making multi-step tasks faster and less confusing

If you adjust your data contribution settings to opt out of metadata and/or in-app data, or delete an app or site, we will:

  • remove the corresponding in-app data within 30 days from our datasets used to improve apps and experiences for all customers, and corresponding content attributes within 90 days

  • retrain any models previously trained on that data

  • stop collecting new metadata and/or in-app data from the corresponding app

  • stop identifying recurring patterns from data you previously contributed

Still need help?

The Atlassian Community is here for you.