NexusAI by MPBusiness

Security built around controlled access and traceable data.

This page describes controls implemented today. It does not claim certifications or enterprise controls NexusAI has not completed.

Authenticated workspaces

Private application routes require an authenticated user and workspace context before business features are available.

Workspace-scoped access

Business data requests are resolved in the signed-in workspace context. Privileged server operations are kept behind server-side routes rather than exposed as public browser credentials.

Source-backed reporting

Reporting is built from persisted imports and validated records. Invalid or excluded values are kept out of KPI calculations rather than silently invented.

Read-only AI Advisor

AI Advisor accepts a bounded, source-backed evidence packet marked read-only. It is instructed not to change records, send messages, approve actions, or perform external actions.

Bounded AI requests

Questions and evidence packets have size limits, requests time out safely, and returned evidence references are restricted to references present in the supplied packet.

Server-side secrets

AI gateway credentials and other privileged credentials are read on the server. They are not intentionally shipped in client-side application code.

Release verification

Production changes use automated checks and deployment verification before they are treated as complete.

Current scope

NexusAI does not currently claim SOC 2 certification, SAML/SSO, SCIM, or customer-managed encryption keys. Those capabilities should only be published after implementation and verification.

For AI-specific handling, see AI data handling. For contractual privacy terms, see Privacy.