Can useful models run where evidence lives?
Sending evidence out to be read widens who can see it. Cyberdroid AI Detection and Cyberdroid AI Investigation run their models in your environment by default.
See the methodMethod
Inference stays in your environment. Nothing leaves unless you turn on a fallback.
Model serving is shared and on premises, with key-based access and per-key concurrency quotas.
Each request is routed to a model by role, and the routing metadata is visible.
AI Detection runs on your hardware, on CPUs, with local models and no GPU.
An optional online fallback is a policy you configure and can see.
Measures
CPU, memory and storage used, measured on your hardware during a proof of value. We publish the method, not a sizing figure.
Limits
Local models trade capability against hardware. A fallback, once you turn it on, is visible in policy.
