Trust and Compliance — ALFRED™
Legal

Trust and Compliance

Effective Date: July 3, 2026  ·  Operated by: Fintech Reimagined LLC

This page explains how ALFRED™ is governed — how the platform uses artificial intelligence, how compliance is enforced across the Entrepreneur, CDFI, and Lender engines, and how every state assignment and file action is logged and auditable.

ALFRED is built on a simple governing principle: the platform’s intelligence carries the logic of the P3 Framework™ and Readiness Intelligence™ — it does not replace the judgment of the human institutions ALFRED serves.

ALFRED’s AI layer interprets, translates, and drafts within bounded system rules. It does not make underwriting judgments, policy exceptions, or autonomous release decisions.

Section 1

How ALFRED Uses AI

ALFRED uses artificial intelligence in specific, bounded ways in support of the P3 Framework™:

  • Generating personalized readiness guidance from structured data in a user’s file — constrained to a defined reason code catalog
  • Evaluating changes in connected banking or verification data to determine whether a Readiness State update is warranted
  • Powering ALFRED Chat, a conversational interface that responds to user questions in ALFRED’s voice and reads live file status

ALFRED does not use AI for:

  • Credit scoring or risk modeling
  • Approval probability estimation
  • Income or revenue verification beyond threshold comparison
  • Any decision that constitutes underwriting judgment

ALFRED Chat is an AI-powered conversational interface, not an autonomous AI agent. It responds to what a user asks — it does not initiate actions or make decisions on a user’s behalf.

Section 2

Readiness State Logic

PAUSE, PREPARE, and PROCEED — ALFRED’s three Readiness States — are produced by a deterministic, rules-based logic engine, not by predictive AI. Readiness State is the governing output across all engine variants; Readiness Index (used in the Entrepreneur engine) scores familiarity and does not itself decide state.

A Complete Verified File is assembled only once a user reaches the Proceed state and explicitly approves submission. No file is transmitted to a lender, CDFI, or capital provider without that approval.

Section 3

CDFI Compliance and Dual-Gate Enforcement

For institutional deployments, ALFRED enforces a dual-gate compliance structure before any capital opportunity is surfaced to a user:

  • Target Market compliance — verifying alignment with a CDFI’s defined Target Market criteria
  • Minimum product eligibility — confirming the user meets a program’s baseline eligibility parameters

Compliance is treated as a product attribute, not a universal gate — logic parameters are scoped to each institution’s specific programs. ALFRED’s engine structure remains constant across institutions; only program logic parameters change.

Section 4

Exception Handling and Safe Mode

ALFRED maintains a defined path for handling disputes, sync failures, contradictory data, or unexpected system behavior:

  • The affected automated action is flagged and frozen pending review
  • Human review is required before the action resumes
  • The outcome is appended to the audit log, and the prior file version is preserved

Any anomaly, unexpected output, unauthorized write attempt, or policy conflict triggers Safe Mode. In Safe Mode, ALFRED may continue to read and log activity, but external sending, CRM sync, and state advancement are disabled until a documented review restores normal operation.

Section 5

Data Isolation and Access Boundaries

  • ALFRED’s system components operate only within a user’s scoped file and engagement context
  • No component may access or infer from data outside a user’s authorized record set
  • Institutional program data is tenant-bound and is not surfaced outside that institution’s implementation
  • Lender identity is not exposed to a user before the user selects a capital opportunity
Section 6

Audit Logging

Every consequential action within ALFRED is logged, and logs are immutable. The audit trail is available to institutional administrators for compliance, reporting, and review purposes.

Action What Is Logged
State assignment Timestamp, reason codes, gate status, and engine version
Email triggered State, template used, dynamic fields populated, and send timestamp
Guidance generated Reason codes used, generation timestamp, and output reference
CRM sync Payload fields, transmission timestamp, and confirmation receipt
State regression Previous state, new state, triggering data change, and timestamp
Failed gate attempt Missing fields, reason codes, and timestamp of blocked submission

Audit logs are retained for a minimum of five years, consistent with CDFI Fund documentation retention requirements, and are available for institutional review or regulatory examination as required.

Section 7

No Financing Guarantee

ALFRED does not guarantee, promise, or imply approval for any financing product, loan, line of credit, or capital instrument. Readiness State, Readiness Index, and the Complete Verified File are indicators of preparation — not commitments to funding. All financing decisions remain the exclusive authority of the relevant lender, CDFI, or capital provider.

Questions or Compliance Inquiries

For questions regarding this Trust and Compliance page, AI governance, audit requests, or institutional compliance matters, contact Fintech Reimagined LLC at:

compliance@getalfredapp.com  ·  hello@getalfredapp.com

ALFRED™ is a trademark of Fintech Reimagined LLC. All rights reserved.