PipeLedger AI

SECURITY & TRUST

The security layer your ERP was never built to have.

The problem

The fear isn't that AI is useless on financial data — it's that one prompt exposes executive comp, an unreleased number, or another entity's books. Accounting systems can't prevent that: they have no row-level security, no column masking, no gate to enforce who sees what on every query. The risk isn't the model — it's the ledger underneath it.

How it works

Six guards around your ledger — always on

Your ledger
all guards, always on
Is it you?
2FA enforced
Trusted place?
IP allowlist
Right key?
audience-bound
Allowed to see?
role · scope · clearance
Person cleared it?
approved snapshot
Sealed pipes
encrypted · HMAC
01
Tenant isolation

Organization-scoped datasets and access checks establish the tenant boundary. Server-owned paths must enforce that organization scope explicitly.

02
Approved snapshots only

Governed delivery resolves eligible, approved published data and rejects retracted, superseded, expired, or unpublished versions.

03
Query-time enforcement

MCP, REST, and CLI use the shared query-policy boundary. Configured BI sources receive separately sanitized published datasets.

Audit trail
Material governance actions recorded as append-only evidence; successful revocations are enforced at the authorization boundary
Encryption
Managed encryption at rest · HTTPS/TLS in transit · current providers listed on Trust
Compliance
SOC 2 in preparation · runs on SOC 2 and ISO 27001 certified infrastructure
Processor role
Privacy roles and commitments are governed by applicable law and executed terms; authoritative transforms are deterministic
Security & Trust | PipeLedger AI