No client files in discovery.
The public intake asks for workflow pressure, systems, approvers, and desired outcomes only.
Security and privacy boundary.
LedgerPilot AI is a discovery and AI-use pilot-scoping system for accounting firms. It helps identify a workflow pressure point, recommend one controlled AI-use lane, and prepare a partner-ready review packet without client files during discovery.
Professionals keep control.
LedgerPilot scopes where AI could support work. It does not make final professional conclusions.
Implementation comes later.
Any production integration should go through vendor, security, privacy, and professional review.
Discovery should avoid private client data.
- Client names
- SSNs, EINs, or tax IDs
- Payroll registers
- Bank data
- W-2s or 1099s
- Tax returns
- Audit evidence
- Private financial statements
Professional judgment stays with the firm.
- Tax advice
- Legal advice
- Audit opinions or assertions
- Payroll judgment
- Accounting treatment decisions
- Client-facing messages ready for release
- Filing positions
- Payment or invoice release authority
What a firm-wide rollout would need.
Before production use inside a firm, LedgerPilot should be deployed with firm-approved hosting, managed database, SSO/OIDC/SAML, role-based access, named-user audit logs, retention and deletion policy, encrypted backups, vendor review, firm-approved source repositories, and retrieval governance with citation enforcement.
Start where the workflow is measurable and review-owned.
Good first candidates include billing/WIP review support, document intake routing, bookkeeping close readiness, tax notice owner/deadline routing, payroll change intake checklists, and audit support/PBC owner tracking.