For on-call, SRE and platform teams
The incident is over.
The postmortem takes a minute.
Afterwatch reads your Slack incident thread and writes a blameless postmortem — timeline, root cause, impact and action items with owners — before anyone has to schedule a meeting about it.
Free for 5 postmortems a month. No card needed. Read-only access.
PagerDuty CRITICAL checkout-api p99 latency > 3s
priya db connections maxed on orders-db (500/500)
arjun deploy #4821 added order history join
arjun rolling back #4821
Checkout latency from unindexed join in deploy #4821
Root cause: new join on orders.customer_id had no index, saturating orders-db connections. Resolved by rollback after ~25 min of checkout impact.
- P0 Add index on orders.customer_id — arjun
- P1 Load-test checkout path before re-ship — arjun
Hours
lost scrolling back through Slack to rebuild the timeline
Days later
is when postmortems usually get written — after the details fade
< 1 min
for Afterwatch to draft it while the thread is still fresh
How it works
01
Run it on the incident thread
Hover any message → More actions → Write postmortem. Or type /postmortem to cover the channel's last 24 hours.
02
Afterwatch reads the whole thread
Alerts, hand-offs, commands, the rollback. It builds a timeline and separates root cause from contributing factors.
03
Review, edit, share
A summary lands in the thread with a link to the full postmortem. Edit it in the browser, copy as Markdown, done.
Written the way good postmortems are
Blameless by default
Describes systems and decisions, never people. The tone your SRE handbook asks for, every time.
Doesn't invent facts
Severity, owners and causes come only from what the thread says. Anything missing is marked “Unknown — confirm with team”.
Action items with owners
P0/P1/P2 follow-ups, assigned to whoever committed to them in the thread.
Works with your alerts
PagerDuty, Opsgenie, Datadog and Grafana alert messages in the thread are read as part of the timeline.
Nothing to install
A Slack app plus a web editor. No agents, no access to your infrastructure, no production credentials.
Private by design
Read-only access to channels you invite it to. Tokens encrypted at rest. No training on your data.
Simple pricing
Per Slack workspace. Start free, upgrade when it earns its keep.
Free
$0
For trying it on real incidents.
- ✓5 postmortems / month
- ✓Slack app + web editor
- ✓Markdown export
Team
Most popular$29 / month per workspace
For teams that run postmortems every week.
- ✓Unlimited postmortems
- ✓Postmortem history
- ✓Team sign-in for shared pages
- ✓Priority model for long threads
Business
$99 / month per workspace
For orgs with a postmortem process to follow.
- ✓Everything in Team
- ✓Your postmortem template
- ✓Confluence & Notion export
- ✓GitHub PR context
Questions
What does Afterwatch need access to?+
Read access to the channels you invite it to, permission to post its summary, and user names so the timeline shows people instead of IDs. It never touches your infrastructure.
Is my incident data used to train AI models?+
No. Threads are sent to our AI model provider only to generate your postmortem, on API terms that exclude training. We store the finished postmortem, not the raw thread.
How accurate is it?+
It writes only what the thread supports and flags gaps instead of guessing. Every draft is labelled AI-generated and is meant to be reviewed by the incident lead before sharing.
Does it work with Microsoft Teams?+
Not yet. You can paste any chat log into the web version today. Teams support is on the roadmap — tell us if you need it.
Can it follow our postmortem template?+
Business plans can upload a template; Afterwatch fills your sections in your order.
Write your next postmortem in a minute.
Install in Slack, invite @Afterwatch to your incident channel, run it on a thread.