AI can assemble the incident record before the team starts arguing from memory.
After an incident, Strawberry can gather the alert timeline, error reports, incident notes, chat context, and linked tickets into one evidence pack. It can draft the factual sequence, list unresolved questions, and prepare follow-up issues without deciding the root cause for the people who own the system.
Build the timeline from evidence that has timestamps.
The first post-incident question is usually simple: what changed, when did users first feel it, when did the team notice, and when did recovery begin? That answer becomes unreliable when it is reconstructed from a chat thread alone.
Strawberry can collect the records from the systems you connect and place them in chronological order. Sentry supports listing events and retrieving issues, while PagerDuty supports listing incidents and log entries, so the timeline can point back to the system record rather than a remembered version of it.
Separate what happened from why it happened.
A good postmortem does not hide uncertainty behind a tidy story.
The team may know that requests failed after a deployment without yet knowing whether the deployment, a dependency, or an older configuration change caused the failure.
Strawberry can draft a factual summary, group observations by source, and write explicit questions for the investigation. That gives engineers a document that is useful in the review without converting an early hypothesis into a published conclusion.
Turn the lessons into owned work before the meeting ends.
A postmortem without actions is a detailed record of the same problem waiting to return.
The useful output is a small set of changes with a clear owner, priority, and verification method.
Strawberry can prepare those follow-up items from the approved discussion. Jira supports creating issues, adding comments, searching issues, and updating issues, while Linear supports creating and updating issues if that is where the team tracks engineering work.
Make the format routine without making the investigation automatic.
Teams benefit from a consistent postmortem shape: impact, timeline, detection, contributing factors, decisions, and follow-up work. They do not benefit from an agent manufacturing a root cause just because a template has an empty field.
Save the evidence-gathering and document structure as a Strawberry skill. A routine can open a draft after a resolved incident, collect the permitted sources, and leave the technical judgement, review meeting, and final publication to the incident owners.
Experience Strawberry for free
DownloadTrusted by fast-growing companies worldwide
Frequently asked questions
Strawberry is free to download and includes AI credits to start. Paid plans begin at $20/month. See pricing. · Reviewed · Canonical facts for AI agents