After an emergency is resolved, the Rescue dashboard can generate an AI-written summary of what happened — a structured, plain-language report covering the timeline, devices involved, people notified, and observations about how the incident was handled. It is built for debriefs, internal documentation, insurance records, and compliance reporting.
The report is generated on demand. It does not produce itself automatically when an emergency closes — an administrator or authorized user initiates it from within the event record.
Who can generate and view a report
Any user with access to the Emergencies section of the dashboard can generate and view an AI after-action report. Access to Emergencies is controlled by user role and custom permissions. If you can see an event in the Emergencies log, you can generate a report for it.
How to generate a report
Reports can be generated for any resolved emergency in your archive, including past events. You are not limited to recent incidents.
What the report covers
The report is structured consistently across all events, but its content depends on what data was available during the incident. A report for an emergency triggered by a physical device will look different from one triggered through the STOPit Notify app — because they generate different types of data.
Narrative summary A plain-language description of how the emergency was declared, by whom or what device, what level of response was triggered, and how and when it was resolved.
Initial locations and movement For Card-triggered emergencies, the report uses map snapshots to document where Card carriers were located at the time of the alert and — if additional snapshots were captured — how positions changed during the incident. For STOPit-originated emergencies and Raft-triggered emergencies, map data may show only fixed infrastructure (Alert Stations, Repeaters, mapped safety assets) rather than personnel locations if no Cards were in range during the event.
Combined timeline A chronological log drawing from the event record and map snapshots — declaration timestamp, system messages, acknowledgements, resolution, and any intermediate updates captured during the incident.
Duration insights Time from trigger to acknowledgement, time from trigger to resolution, and any notable gaps — for example, a long period between declaration and the first logged action.
Key observations The AI flags patterns in the event data that may be relevant for review: whether the same person declared and resolved the emergency, whether no acknowledgement was recorded, whether intermediate status updates were absent, or whether device location labels were unclear.
Recommendations Actionable suggestions based on what the report found. These are generated from the specific data of that event — not generic advice. Common recommendations include standardizing acknowledgement practices, adding in-chat status updates during incidents, improving device or location labeling, and capturing additional map snapshots for high-severity events.
How reports differ by emergency type
The quality and completeness of an AI after-action report depends on how much data was captured during the incident.
Emergencies triggered by a Rescue Card or Raft: These generate the richest reports. The triggering Card is identified by name, the map captures device locations at the moment of the alert, nearby devices and their positions are visible, and the device's signal path through the Repeater mesh is logged. The report can speak to who was near the incident, where they were, and how quickly the event was closed.
Emergencies declared through STOPit Notify: These emergencies originate in the Notify app and are reported in Rescue as their own events. The report captures who declared the emergency, the category, timestamps, and resolution. Map data reflects fixed assets only unless Cards were active and in range during the event. The report may flag gaps, such as no acknowledgement recorded or no intermediate updates, based on what was logged in the event record.
What makes a report more complete
The AI report can only analyze data that was captured during the incident. A few practices consistently produce more useful reports:
Exporting and sharing
Click Export as PDF within the event record to download a formatted copy of the report. The PDF includes the full report — narrative, timeline, observations, and recommendations. It is suitable for board presentations, insurance submissions, regulatory filings, and internal after-action reviews.
The PDF reflects the report at the time it was generated. If you re-generate the report later, the content may differ if the underlying event data has been updated.
Generating a report does not modify the event record
Generating or exporting an AI after-action report does not change anything in the Emergencies log. The original event record, its timestamps, and its logged data remain exactly as captured.
The AI after-action report lives inside individual event records on the Emergencies page. It does not appear on the Home page, Devices page, or anywhere else in the dashboard.