Privacy Policy

12. Service Quality Review for Jira Service Management

SQR coverage was published on October 22, 2026 and applies from the product’s first public availability through Atlassian Marketplace. Publication of this policy does not announce SQR’s Marketplace availability. Each customer’s agreement takes effect under its applicable agreement.

Who we are and what this policy covers

Service Quality Review for Jira Service Management ("the app") is an app for Jira Service Management, provided by Magnascale Inc., Suite 411, 4033 Hurontario St, Mississauga, ON L4Z 0B8, Canada ("we", "us"). This policy explains what information the app handles and what we do with information we receive about our customers.

There are two kinds of information:

• Customer data in the app. This is what your organization's users create or reference while using the app in your Atlassian site: review programs, reviews, cited ticket excerpts, account identifiers and similar. Your organization decides how the app is used and is responsible for that data. We process it only to provide the app, on your organization's behalf. For customer-controlled data, Magnascale acts as processor or service provider to the extent applicable law recognizes that role; your organization is controller or otherwise determines the processing purposes. The shared DPA includes the SQR processing schedule.
• Information about you as our customer. This includes licensing and billing contact details Atlassian shares with us for your subscription, and the contact details and correspondence used to manage your support request. We use it to provide the subscription and support, and to meet legal obligations. Magnascale acts as an independent controller for its own customer/account records. The applicable bases include providing the contracted service, responding to support requests, legitimate security/business administration and legal obligations, as applicable to the actual records received. The SQR providers and support-records subsection describes the contact providers, Marketplace data categories and vendor-record practices. Customer-controlled diagnostic log data remains subject to customer instructions and the shared DPA when handled for troubleshooting.

Where the app runs

The app is built on Atlassian Forge. It runs on Atlassian's infrastructure, and persistent runtime records are stored in Atlassian’s Forge Key-Value Store and Custom Entity Store associated with your Atlassian site. The browser keeps account and record identifiers, preferences and progress/recovery state. Session storage can include pending commands, selected reviewer identifiers and manager-entered recovery reasons; those reasons can contain personal information. Browser-held records follow the browser’s storage lifecycle and can be cleared through its controls. Downloaded CSVs stay on the recipient’s device, and confirmed corrective work is stored in Jira; app deletion does not recall those delivered records. The app runtime has no configured external egress or vendor-hosted backend. It uses Atlassian APIs and Atlassian-hosted Forge LLMs. We do not run analytics or tracking in the app. Support messages and subscription contacts are separate vendor records; the contact providers and vendor-record practices are described in the SQR providers and support-records subsection.

What the app stores

• Atlassian account IDs of people who use or appear in the app: administrators, managers, reviewers, the agents whose tickets are reviewed, and authors of ticket comments that a reviewer cites. Structured profile names, emails and pictures are looked up live rather than stored as a directory. Names or emails can still occur in ticket excerpts, AI inputs and user-written text.
• References to Jira content: project, ticket and comment identifiers, timestamps, and fingerprints used to detect whether cited text has changed.
• Cited excerpts: the exact passages from a ticket that a reviewer (or an AI suggestion the reviewer requested) quotes as evidence. Submitted review evidence keeps cited excerpts rather than a general ticket mirror. Temporary AI input can include larger bounded selections as described below. Excerpts can contain whatever personal information the ticket text contains.
• Review content your users create: judgments, ratings, notes, drafts, corrections, calibration results, coaching summaries, findings and corrective work descriptions.
• Configuration: program names, review criteria, people lists, time zones and schedule settings.
• AI records, if your users use AI suggestions: the ticket evidence sent to the model, the validated suggestions, and usage records (see SQR AI suggestions).
• Operational records: in-app updates, audit history and processing status.

The app writes to Jira only when a manager explicitly confirms a corrective work item. It creates that item as the manager, using the text the manager wrote.

How the information is used

Only to provide the app's features to your organization: selecting and assigning reviews, showing evidence, saving and reporting reviews, calibration, findings, reminders, and AI suggestions when requested. The app does not sell customer data, use it for advertising or run model training. Atlassian provides inference under its platform terms; this is not a separate guarantee about the platform provider's practices.

Access within your organization follows your settings and Jira permissions. People see review content only if they hold the right role in that review program and can currently see the underlying Jira ticket and comments. The reviewed-agent role alone grants no access to underlying reviews. A manager can deliberately release a separate coaching summary. Source-free report numbers and completed-results CSV use current Jira issue visibility under their separate numeric-reporting rules and exclude notes, quotations and AI text.

The app is designed to support quality improvement and coaching. It should not be used as the sole basis for automated employment decisions.

AI suggestions

AI suggestions are optional. They run only when a reviewer asks for them on a ticket assigned to that reviewer. The app then sends the review criteria and a limited selection of that ticket's text, which the reviewer is allowed to see, to Forge LLMs: large language models hosted by Atlassian. Atlassian applies its own content moderation to these requests.

Citations in observations and inferences are validated against supplied evidence; an insufficient-evidence result can contain no quotation. Citation validation does not guarantee a correct interpretation. They are labelled as suggestions, and a person makes every judgment. AI does not submit reviews or write to Jira. A reviewer can apply a suggestion to their draft, edit it and make the final submission. Unused AI material expires after at most 30 days, subject to earlier parent-record expiry. Expired material is withheld from app views; physical cleanup runs asynchronously. Forge LLM processing is provided by Atlassian under its own terms; see Atlassian's Forge LLMs documentation.

How long the app keeps information

The app follows a fixed retention schedule. Expired information is withheld on the next authorization/read check; content already delivered to a browser or downloaded file cannot be recalled. Physical cleanup runs in bounded background passes and can wait while related work is active. The periods below are maximum visibility/retention-policy bounds, not guarantees of physical erasure at the exact deadline.

If your licence ends, the app continues authorized reads, supported exports, privacy actions and retention while the installation remains accessible; new reviews and AI suggestions stop. Atlassian separately controls installation access, suspension and its platform lifecycle.

Deletion and closed accounts

• Closed Atlassian accounts. The implemented account-closure process regularly checks with Atlassian which stored accounts have been closed. Within 30 days of learning that an account is closed, it replaces that person's identity with "Former user" everywhere. It deletes notes, quotations and other free text about that person on reviews of their work, and removes calibration results they took part in. Scores and counts remain under a replacement identifier to preserve history. This is not a guarantee of anonymity in their wider context. Names mentioned in free text on reviews of other people are not automatically searched.
• Deleting data on request. Program managers can permanently delete a program from its settings. Jira or app administrators can delete one program or erase all of the app's data for the site. The app shows how many records will be deleted and asks for confirmation. Deleted data disappears from the app immediately and is removed by background cleanup, after which a content-free receipt is kept for 30 days. A program deletion also keeps the indefinite content-free closure marker described above to refuse late writes to that deleted program. Managers can also delete review periods that have not started. For help, contact privacy@magnascale.com.
• Uninstalling. After uninstall, Forge-hosted runtime data follows Atlassian’s then-current soft-delete, recovery and backup lifecycle. A reinstall starts as a new installation; a request to reconnect previous data must be made within 21 days of uninstall through Atlassian’s recovery process. That is a recovery-request window, not a maximum physical-deletion period. See Atlassian’s hosted-storage lifecycle. The runtime app has no external copy store. Separate vendor support records and temporary troubleshooting copies follow the vendor-record/log practices below; customer downloads and Jira work items remain under customer/recipient control.
• Backups. Deletion in the app does not immediately remove copies held in Atlassian's platform backups. Those follow Atlassian's own schedules.

Genuine closed-account erasure and full-site erasure have not been verified in native platform acceptance. Retention cleanup completion under ongoing activity remains a post-launch validation item. These evidence limits do not change the fixed retention schedule or create a physical-erasure deadline.

Providers, support records and sharing

Atlassian provides the runtime, hosted storage, queues, platform logs, Jira/JSM APIs and Forge LLM processing used by the app. The app has no separate AI service, external analytics or Magnascale-hosted application backend. Atlassian supplies the disclosed platform processing; providers that act as processors are identified under the shared DPA and its applicable service-provider disclosures. No app-configured external egress does not mean there are no platform or support providers. Any legally required disclosure or business transfer follows the existing shared policy and applicable law; the shared policy applies to those disclosures.

The existing contact form receives name, email, topic and message. Cloudflare provides website hosting/protection and Turnstile verification; Amazon SES delivers the notification to a fixed private business destination hosted by Google Workspace, using the visitor’s address as Reply-To. The website does not maintain a contact-submission database or intentionally log message content or credentials. Cloudflare processes technical request/browser data, and the handler uses a keyed address-derived identifier for rate limiting. The current shared policy excludes website analytics, advertising and session-recording scripts.

Contact correspondence and subscription/business records follow the shared Magnascale policy: retain only as necessary for the inquiry or support relationship, subscription administration, legitimate business records and applicable legal obligations. This does not promise a fixed mailbox deletion interval. Atlassian makes customer contact details (including email), licensing/evaluation/subscription and transaction records available for paid apps; we use information provided to us for subscription administration, billing, support and legal obligations. This does not establish a separate CRM or routine report-export store.

Magnascale does not maintain a routine archive of customer logs. We process customer-controlled platform log data under customer instructions and the shared DPA only when needed for troubleshooting. If we temporarily download customer logs, we delete those copies promptly when troubleshooting is complete. We may retain the minimal findings needed to fix a bug or improve the product, without raw log payloads or unnecessary customer identifiers. Atlassian manages platform logs and backups under its own lifecycle; availability through its developer console is not a physical-erasure deadline. Do not send ticket bodies, credentials, tokens or private review content through support. SQR customer support is handled by the Magnascale owner and the documented providers, with no additional support tools or recipients. See the SQR DPA schedule.

Security

The app runs on Atlassian Forge, which provides the hosting, storage encryption and network security. In the app, every request is checked on the server against the user's program role and current Jira permissions. App administrators do not automatically get access to review content. Logs contain operation codes and opaque identifiers, not ticket text, names or email addresses. Applicable Marketplace vulnerability and incident obligations are separate from support response targets. Report security issues to security@magnascale.com.

Magnascale’s owner confirms that applicable online and offline data and copies, including temporary customer troubleshooting material, are encrypted. This describes an owner-confirmed operating practice, not an independently audited certification or verification of provider, account, device, or encryption-key settings. Temporary troubleshooting copies are processed and stored while they are in use.

International transfers

Persistent app data follows the host product's supported Forge storage residency pin and migration. This does not guarantee the location of every invocation, AI request, log or support interaction. International transfers use the shared DPA's conditional framework: a valid adequacy mechanism where applicable, otherwise the applicable SCCs, UK addendum or Swiss adaptations where required. The current shared policy identifies Canadian operational access for Conf2Git and possible processing outside Canada by Cloudflare, AWS and Google. The existing contact-mail setup uses SES US East (Ohio); the website Worker uses Cloudflare’s distributed Default placement, rather than a Canada-only location. The Magnascale owner operates from Canada. Other provider processing locations follow the applicable provider services and shared transfer framework; no specific Google mailbox region or exhaustive country inventory is asserted. This framework does not promise Canada-only or EEA-only processing.

Your rights

If your organization uses the app, it controls the data about you in the app. Please contact your organization's Jira administrator first. We will help your organization respond to requests to access, correct or delete personal data. For information we hold about you as a customer contact, write to privacy@magnascale.com. Depending on applicable law, you may also have rights to restrict or object to processing, obtain portability, withdraw consent where consent is relied on, and complain to the competent privacy regulator. Requests about customer-controlled app data go to your organization; Magnascale assists it under the shared DPA and applicable response requirements. Vendor account/support records follow the shared necessity-based retention rule and the applicable legal bases.

Children

SQR is a business tool and is not directed at children.