Privacy Protocol

Revision 2.1 • Effective Feb 2026

"Your architectural secrets are your own. We only provide the shrink-wrap."

Neural Safety

All model weights uploaded to EdgeAI are processed in isolated encrypted containers. We do not store your original uncompressed weights after the session ends unless explicitly archived by the operator.

Data Observability

We track metadata (layer counts, parameter sizes, accuracy metrics) to improve our compression algorithms. No actual weight values or sensitive architecture details are ever transmitted to third-party providers.

Zero-Knowledge Local Ingest

For enterprise users, we offer a local-only CLI tool that performs 100% of the compression logic on your own premises, only syncing billing metadata with our core.

Security Disclosures

EdgeAI employs SOC2 Type II compliant infrastructure. All neural data at rest is encrypted via AES-256-GCM. Connections are secured using TLS 1.3. For security audits, please contact our Legal Neural Node.