Separate facts from conclusions
Perception produces physical events such as a person entering a zone. Correlation and risk remain separate layers, and neither declares guilt or intent.
Responsible Visual Intelligence
Evidence should support judgment, not replace it.
SightGuard is designed to report observable activity, explain why a situation was flagged, and keep consequential decisions with authorized people.
Keep humans accountable
Investigators review evidence and record explicit outcomes. Automated scores remain signals, not determinations.
Preserve the history
Detector confidence, model version, correlation rule, risk score, evidence, reviewer, notes, and timestamps remain traceable.
Privacy architecture
Move less video. Keep more context.
The intended SightGuard Edge architecture processes continuous video locally. Cloud services receive metadata, selected evidence, and configured uploads rather than requiring continuous raw CCTV transfer.
Be honest about capability
Computer-vision performance depends on camera placement, lighting, occlusion, scene configuration, and representative evaluation data. SightGuard will not present a detector as reliable without measured results and documented limitations.
Use neutral operational language
The system uses terms such as exception, anomaly, requires review, and potential issue. Human outcomes may include confirmed loss, policy violation, legitimate activity, false positive, or needs investigation.
This page describes SightGuard's product principles during development. Detailed customer controls, retention settings, and contractual commitments will be documented as the platform matures.
Work with us
Help us test these principles against real operations.
Our Early Access Program is intended to evaluate useful workflows, limitations, and responsible deployment requirements.