← Atlassian Interview InsightsSoftware Engineer

Atlassian·Software Engineer·Onsite - System Design / Architecture·Intermediate

IntermediatePrefer not to say
Jun 2026

Summary

Atlassian SWE interview that was basically one big OOD problem about a CI/CD release notification service. The question had enough moving parts that it felt more like a mini system design than a pure coding round, which I wasn't totally expecting.

Questions Asked (1)

Q1

Design an object-oriented model for a CI/CD release notification service. The service runs daily deployments and notifies each commit author when their change is successfully deployed. If a deployment fails and rolls forward to the next version, authors from the failed deployment must be carried over and included in the next version's notifications. Design the core classes (Deployment, Version, Author, Notifier), the rollover logic, and the notify() API.

System DesignData ModelingAPI & Integrations
Author's notes

I started by sketching out the classes and immediately got tangled on the rollover part.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying requirements and defining the core classes with their responsibilities and relationships. Then walk through the rollover logic and notify() API, emphasizing how authors are carried over and how notifications are triggered. Finally, discuss edge cases, extensibility, and potential improvements.

Pro tip: Demonstrate awareness of real-world concerns like idempotency, failure handling, and scalability by mentioning how you'd avoid duplicate notifications and handle large deployments.

1. Clarify requirements and scope

Ask clarifying questions about deployment frequency, notification channels, failure handling, and whether authors should be notified per deployment or per version. Confirm assumptions about rollover behavior.

2. Define core classes and relationships

Identify Deployment, Version, Author, and Notifier classes with their key attributes and methods. Establish associations: a Deployment has a Version and a list of Authors; a Notifier sends notifications to Authors.

3. Design rollover logic

Explain how a failed deployment's authors are carried over to the next version. This could involve a pendingAuthors list on the Version or a separate RolloverManager that aggregates authors across deployments.

4. Design notify() API

Define notify() to accept a Deployment (or Version) and trigger notifications to all associated authors, including carried-over ones. Discuss synchronous vs asynchronous notification and error handling.

5. Discuss edge cases and extensibility

Address scenarios like multiple consecutive failures, duplicate notifications, and scalability. Mention how to extend for different notification channels or retry mechanisms.

Key Points to Mention

  • Clear separation of concerns: Deployment orchestrates, Version holds metadata, Author represents recipients, Notifier handles delivery.
  • Rollover mechanism: maintain a list of pending authors on Version or use a separate component to merge authors from failed deployments.
  • Idempotent notifications: ensure authors are not notified multiple times for the same version, possibly using a notified flag or set.
  • Notification API design: notify() should be flexible, perhaps accepting a Notifier interface for different channels (email, Slack).
  • Handling multiple failures: authors accumulate across failed deployments until a successful deployment notifies them all.
  • Scalability and reliability: consider asynchronous processing, retries, and logging for notifications.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.