00:00:00/Check
Weapon Alert: seconds matter
Weapon Alert is a Svid check that raises a critical alert when a weapon is visible on camera. It goes out by email and JSON webhook at once, with a plain description and a link to the exact moment, so whoever needs to act gets the earliest possible start.
What it does
Svid already describes every frame of the footage it watches. Weapon Alert is a paid check over those descriptions that watches specifically for a visible weapon, a knife or a gun in shot, and treats it as critical the moment it appears: the alert goes straight out by email, with a JSON webhook firing at the same instant. The boundary is honest and worth stating plainly: it can only flag what's visible to a camera. It makes no claims about anything concealed.
Where it applies
Weapon Alert sits in the Security & loss prevention group in the Store, but it isn't built for one setting. Schools, venues, late-night retail, reception desks, transport hubs, it's the same check with the same behaviour on any camera it's enabled on.
What the alert looks like
An alert reads plainly: "Person with a knife visible near the main entrance, 14:03", with a link straight to that moment, on screen and by email. The JSON webhook carries the same event for anything that should happen automatically, a message to a response group, a signal into an existing security system. The design intent is a human with good information, quickly. It's a flag for people, not an automated response.
Where it runs
Like every check, Weapon Alert is switched on per camera, and entrances and public-facing cameras are the usual choice. It works on RTSP feeds, uploaded files, video URLs and a live phone camera. Recorded footage runs through the same pipeline as live, which is mainly useful for testing it against your own cameras before relying on it.
What it costs
The Store prices Weapon Alert from your own usage before you buy, shown as ‘+$X/mo at your usage’. It's added to a monthly subscription priced from your own usage with a $9 a month floor, and the whole derivation is on screen at signup. The pricing section shows the arithmetic. No discount theatre here, just your usage, plainly priced.
- What counts as a weapon?
- A weapon visible on camera, a knife or a gun in view. It flags what can be seen and makes no claims about anything concealed.
- Who gets the alert and how fast?
- Whoever you configure, by email and JSON webhook simultaneously. The webhook can feed an existing security process the moment the alert fires.
- What should happen when an alert fires?
- That's your response plan, and it stays yours. The check's job is to start it earlier, with a description and the exact moment on camera.
Test it against your own cameras
Create an account at app.svid.ai and point Weapon Alert at an entrance camera. If you'd rather talk through where it sits in a response plan first, the calendar link is there too.
Related footage
Solution
Nursery CCTV monitoring without watching the children
How nurseries use existing cameras to flag a child alone where they shouldn't be and answer parents' questions calmly, with no facial recognition and nothing tracked.
Guide
How to get alerts from CCTV footage
To get useful CCTV alerts, have AI read the footage and flag the events you care about: on screen, by email, or by JSON webhook into systems you already run.
Check
Child Safety: live awareness around children
Child Safety is a Svid check that alerts staff the moment a child leaves a supervised area alone, with no facial recognition and no biometric IDs.
Check
Child Safety Review: calm answers after the fact
Child Safety Review is a Svid report check that reconstructs a nursery incident from existing cameras in minutes, so a parent's question gets an accurate answer.
Check
Restricted Zone Intrusion: who's where they shouldn't be
Restricted Zone Intrusion alerts when a person shows up somewhere they shouldn't be, on cameras you already have, with rules by camera and by schedule.