Three products · One stack

The trust stack for governed AI.

Three architecturally distinct layers. Three patent-pending products. Ten UK patent applications pending at UKIPO.

Identity Layer · Safe
synapse-ID

“Your data. Your jurisdiction. Your governance framework applies.”

Identity protection for the AI age. Synapse-ID™ reduces the digital identity footprint individuals leave across email, banking, browsing, and AI interactions. Privacy by architecture. Data sovereignty by design.

8 UKIPO patent applications / Identity Footprint Reduction / Privacy by Architecture
Explore Synapse-ID
Governance Layer · Reliable
aiGUARD

Thames Sentinel™ · Non-bypassable execution-control architecture

A runtime governance layer designed to support EU AI Act risk-management, logging, and transparency obligations. Thames Sentinel™ (GB2603184.9) evaluates confidence, consequence, and user state at runtime, per inference. Architecturally enforced, the response generation component is structurally prevented from initiating delivery without a valid execution permission.

GB2603184.9 / Filed 11 February 2026 / IBM Reference Build
Explore aiGUARD
Evidence Layer · Accountable
aiGEC

The proof that your AI was governed. Per inference.

The Governance Execution Certificate. A cryptographically verifiable, per-inference certificate proving governance was applied to a specific output. Issued only upon successful completion of runtime execution-permission evaluation. Signed by hardware-protected keys. Externally verifiable without internal system access. Architecturally unforgeable.

GB2607087.0 / Filed 29 March 2026 / Per-inference cryptographic proof
Explore aiGEC