Post-mortems
Post-mortems and incident reports, drafted from the incident itself
Oncallz turns an incident into a post-mortem draft: title, severity, a timeline built from the incident’s own events and any root cause already recorded.
Your team adds impact and action items with owners and due dates, publishes it internally, and can publish a separate plain-language statement for customers.
Starts from the record
The timeline is built from real incident events, with flapping alerts folded, so nobody has to reconstruct it from chat history.
Action items with owners
Each item has an assignee and a due date, and open items stay on the dashboard and in the daily digest until they are done.
Customer statement, kept separate
Customers only ever see the public statement you write. Nothing from the internal record reaches the public page.
What goes into an Oncallz post-mortem?
A post-mortem belongs to one incident and has the parts below. Next to the document, measured facts come straight from the incident and cannot be edited: duration, start and recovery time, team, source, how many times the alert fired, time to acknowledge and time to resolve.
- Title, impact type (internal, external or both) and outage severity.
- Impact — one line per thing that broke: who or what was affected, and for how long.
- Timeline — a chronological account of detection, response and recovery.
- Root cause and resolution.
- Action items — what needs to happen, who owns it and by when.
- Public statement — the only part customers ever see.
How is the draft generated?
Opening a post-mortem pre-fills the fields the incident actually knows: the title, the severity, a timeline built from its events in your timezone, and the root cause and remediation from a root cause analysis (RCA) if a responder already recorded one.
Notification deliveries are left out of the timeline, and a re-triggering alert is folded into one line so it does not bury everything else.
Anything that is a judgement — who was affected, whether the impact was internal or external, what should change — is left blank on purpose. Nothing is stored until you save, and “fill the blanks from the incident” only fills empty fields, so it never overwrites what you have written.
How do action items get followed up?
Action items are where post-mortems usually die, so Oncallz keeps them visible after the document is published. A post-mortem counts as outstanding until it is published and every action item is ticked off.
Outstanding reviews and your own open items, soonest due first, appear on the dashboard and in the mobile app, and the daily digest reports on them too.
How do public incident reports work?
When an outage affected customers, write a short public statement in the post-mortem: what customers experienced and what you did about it, in their words. Preview it exactly as a reader will see it, then publish.
The report gets its own page, with no login, showing your organization name, the statement, when the outage started and ended, and how long it lasted.
Nothing else from the post-mortem reaches that page: not the internal title (often a raw alert name), the timeline, root cause, action items, team or alert source. You can take the page down at any time, and readers can switch it between English and Turkish.
Post-mortem or RCA — which one do I need?
Not every incident deserves a full review, so Oncallz has both. A lightweight root cause analysis — what happened, root cause, remediation — can be recorded on any incident, or the incident can be marked as not needing one. A post-mortem is for significant outages.
If an RCA exists, its root cause and remediation carry over into the post-mortem draft, so nobody writes the same paragraph twice.
Frequently asked questions
Does Oncallz write the post-mortem with AI?
No. The draft is assembled from facts the incident already holds — its events, times, severity and any RCA — rather than generated by a model, so it contains nothing that did not happen. Judgement fields such as impact and action items are left for your team.
The AI features in Oncallz are separate and run on your own provider key.
Are Oncallz post-mortems blameless?
The structure supports a blameless review: the timeline records what happened and when, and the fields ask about impact, cause and resolution rather than who made a mistake. Whether a review is blameless ultimately depends on how your team writes it.
The timeline does name who acknowledged, escalated or commented, because that is part of the factual record.
Who can see a post-mortem?
Post-mortems live inside your organization and are visible to signed-in members who can see the incident itself; publishing to the team marks the write-up as final.
Customers see nothing unless you separately publish the public statement, which appears on its own page without the internal title, timeline, root cause, action items or team.
Can I unpublish a public incident report?
Yes. Taking the report down makes the public link stop working straight away, and you can edit the statement and publish it again later. Publishing requires a public statement to be written first, and a confirmation step reminds you that anyone with the link will be able to read it.
Are post-mortems included in the plan?
Yes. Post-mortems, RCAs, action items and public incident reports are all part of the single Oncallz plan at $4 per user per month, billed quarterly. There is no separate incident-management tier. Readers of a public incident report do not need an account and do not count as users.
Related
- Status page A public status page with components, 90-day uptime, incident updates and email or Slack subscribers. Alerts prepare drafts; people publish them.
- AI investigation Before it pages, Oncallz summarises the alert and suggests a probable cause and next step; optionally it skips the phone call for clear noise. Your own AI key.
- Escalation policies Multi-step escalation that climbs from the on-call engineer to the team and then leads, paging until someone acknowledges or the policy runs out.
Ready when the pager goes off.
Create your workspace in minutes. Free to start, no credit card — $4 per user a month when you’re ready.