Don't Just Get Alerted, Let the Edge Take Action
Unlike camera AI that only sends an alert, Grablo Reflex reads your cameras, microphones and sensors on the controller itself. It acts on what it finds, from a spoken warning to switching a relay, with no cloud in the loop.
Source, rule, action. That's the whole model.
There is nothing to script and nothing to deploy. You pick what to watch, describe when it counts, and choose what should happen.
Connect a source
Point Reflex at an RTSP camera, a microphone, a Modbus register or a GPIO pin. The on-device AI starts analysing straight away.
Describe the moment
Combine conditions by clicking: a known face, a specific sound, a value crossing a threshold, a time window. There is no expression language to learn.
Let it fire
Actions run on the controller: switch a relay, send a push with a snapshot, speak a warning, write a log row, publish to MQTT.
Not just cameras.
Reflex started as a camera product. It isn't one any more. Every input below drives the same rules and the same actions.
What people actually run.
Each of these is a few clicks: a condition on one side, actions on the other.
Give your controller reflexes.
Free to start. Bring one camera or one sensor and wire your first reaction in minutes.