Incident Time Machine

From incident to insight in seconds — engineering teams move at race speed.

Comment

Inspiration

In high-performance environments like Formula 1 racing, every decision is logged, analyzed, and replayed to improve future outcomes. After every race, teams study telemetry data to understand what went wrong, what worked, and how to gain milliseconds next time.

We noticed a strong parallel in software engineering — incidents are our races, but post-incident analysis is often fragmented across Jira comments, logs, and timelines. Teams lose valuable context, learning is manual, and insights are easily missed. This inspired us to build a way to replay incidents like a race replay, turning chaos into clarity.

What it does

Incident Time Machine is a Jira-native app that reconstructs incidents into a clear, chronological timeline. It automatically:

  • Builds a structured incident timeline from Jira issue activity
  • Generates an AI-powered narrative summary of what happened
  • Highlights key phases like detection, mitigation, and resolution
  • Suggests relevant response playbooks for faster learning and recovery

The result is a single, readable “replay” of an incident — making post-mortems faster, clearer, and more actionable.

How we built it

The app is built using Atlassian Forge, ensuring it runs securely and natively inside Jira.

Key components include:

  • Forge UI for the incident timeline and replay interface
  • Jira APIs to extract issue events, comments, and metadata
  • AI logic to transform structured events into human-readable narratives
  • Timeline grouping logic to surface critical moments and decisions

The architecture follows Atlassian best practices and is designed to be scalable, secure, and extensible.

Challenges we ran into

One major challenge was making raw Jira data meaningful. Jira events are detailed but not inherently narrative, so translating them into a clear story required careful structuring and prioritization.

Another challenge was balancing automation with trust — ensuring AI summaries remain accurate, transparent, and helpful rather than overwhelming or misleading.

Accomplishments that we're proud of

  • Built a fully functional Atlassian Forge app from scratch during the hackathon.
  • Successfully integrated Jira workflows with an AI-powered assistant to speed up incident analysis.
  • Designed the solution around real-world racing scenarios, aligning engineering speed with race-day decision making.
  • Delivered a complete end-to-end demo, including UI, logic, and a clear product narrative.

What we learned

We learned that incident response isn’t just about fixing problems — it’s about learning from them. Presenting data as a story dramatically improves understanding, collaboration, and retention across teams.

We also gained deep hands-on experience building production-ready apps with Forge and designing AI features that enhance, rather than replace, human decision-making.

What's next for Incident Time Machine

Next, we plan to:

  • Add cross-incident analytics to detect recurring failure patterns
  • Support multiple Jira projects and incident types
  • Enable exportable post-incident reports
  • Enhance AI recommendations with team-specific playbooks

Just like racing teams continuously refine performance, Incident Time Machine aims to help software teams learn faster and respond smarter — one incident replay at a time.

Built With

Share this project:

Updates