Trust & Security
Built for the technical due-diligence call.
Every customer gets their own database, their own storage, and their own scoped credentials, so the tenant boundary is a property of the architecture rather than of every query anyone ever writes. Automated architecture tests enforce our core tenant-isolation invariants before any change ships, and destructive database operations are held for your confirmation before they run. Our contractual commitments are the ones in the Master Services Agreement and Data Processing Addendum.
Security details
Architecture, security, subprocessors, AI safety, and disclosure.
Architecture
Dedicated data for every customer: its own database and file storage, with credentials in a managed, tenant-scoped secrets vault. Automated tests enforce our core isolation invariants before any change ships.
Security practices
The concrete controls: encryption in transit and at rest, required MFA and Argon2id password hashing, a managed secrets vault, point-in-time backups, and security invariants enforced in CI on every change.
Subprocessors
The third-party services that process tenant data on Inscendo's behalf: what each one does, where it processes data, and what restrictions Inscendo holds them to.
AI safety
Hallucination risk. Prompt injection. Autonomous-agent governance. The Companion's sensor limits. And a clear list of what Inscendo is NOT a fit for today.
Vulnerability disclosure
How to report a security issue, what's in scope, the safe-harbor terms, and our acknowledgement / triage / disclosure timelines.
Plain-language commitments
These show up on the marketing site, in the contracts, and in the codebase enforcement tests. We don't ship anything that contradicts them.
- No AI training on your data. Customer data is not used to train, fine-tune, or evaluate AI models. We contractually require our AI provider to honor this.
- Tenant-scoped credential storage. Every integration credential lives in a managed secrets vault, scoped to your tenant, and is supplied to the agent at the moment of a call rather than stored in your capsule source. Automated redaction strips recognizable secrets from our operational logs.
- Confirmation on destructive operations. Dropping or truncating a table, and data changes that would cascade into your other automations, are held for your confirmation. The design reduces cross-tenant risk rather than claiming to eliminate it; the MSA is the authority on what we warrant.
- Isolation invariants tested in CI. Automated architecture tests run before any change ships and enforce our core tenant-isolation invariants.
- 30-day log retention. Operational logs are kept for 30 days and credentials we can recognize are stripped from them. Personal details such as names, emails, phone numbers, and addresses are retained as written, so treat agent inputs the way you would any other system of record.
- U.S. only, enforced at the edge. The Service is offered to U.S. businesses, and access from the EEA, UK, Switzerland, China, Russia, and U.S.-sanctioned regions is blocked at the edge rather than left to policy.