What CFSI Loan Management Does
CFSI Loan Management provides construction loan risk management to banks, credit unions, private lenders and family offices: contractor review, project feasibility, fund control, inspections and audits. We use AI-enabled tools to deliver that work faster and more consistently. People remain responsible for every result.
AI may assist with the work. A qualified CFSI professional approves the work. CFSI is accountable for the work.
How we use AI
AI assists; it does not decide. AI-enabled tools help us review documents, research, validate data, identify exceptions, prepare reports and run inspection workflow.
A qualified person approves every deliverable. No report leaves CFSI until a qualified CFSI reviewer has read it in full, checked it against the source documents, personally resolved every issue the system flagged, and approved its release. The reviewer is named on the report. Our quality-control team re-checks a random sample of released reports every month.
We recommend; our clients decide. AI does not make lending, underwriting, credit, loan approval or construction funding decisions. Those decisions belong to our clients. CFSI does not hold or release client funds.
Approved, business-grade tools only. We use AI only through tools that have been evaluated and approved, under business license’s with individual user accounts and terms that prohibit training on our data. Personal and consumer AI accounts may not be used for CFSI work. Every tool is recorded in an internal register and reviewed at least once a year.
Real photos from real sites. Inspection photos are captured on site through our inspection app, with time and location verification. AI-generated or AI-altered images are prohibited as inspection evidence.
How We Handle Client & Borrower Data
We evaluate projects, not borrowers. Our work is based on the project: plans, specifications, budgets, draw requests, site conditions and contractor information. We do not collect or use borrower Social Security numbers, credit reports, bank account numbers or personal financial statements. Information we receive is classified and handled under our internal data classification standard, and the most sensitive class — anything that could identify an individual borrower or owner — carries the tightest handling rules.
Client data is not used to train AI. Client data is used only to perform the services the client engaged us for. Neither CFSI nor our technology providers may use client data to train or fine-tune AI models without the client’s written consent. No client data was used to build or test our platform: development and testing use fictional data only.
Each client’s data stays with that client. Each lender’s information is kept separate. It is not used to serve, score or benchmark any other client. We do not create anonymized or pooled data from client information without the client’s written consent.
Where client data lives.
All client data processed through CFSI’s delivery platform and systems of record is hosted, stored and backed up exclusively in data centers located in the United States.
Access is controlled. Access to systems holding client data is limited to named individual accounts, protected by multi-factor authentication, granted on a least-privilege basis, logged, and reviewed quarterly. The same controls apply to every person and every provider with production or administrative access.
Independent assurance
CFSI’s delivery platform is covered by a SOC 2 Type II performed by Sensiba LLP; the system description covers the Trust Services Criteria for Security, Availability and Confidentiality; no exceptions were noted in the tests of controls. The platform is independently penetration-tested at least annually. Our productivity environment runs on a major cloud suite that maintains its own independent SOC reports. Reports and further detail are available to clients under confidentiality.
Governance and oversight
CFSI’s use of AI is governed by a written AI Use Policy owned by our President. The program includes a tool register, staff and reviewer training, quarterly checks and an annual review. Clients receive advance notice of material changes to the AI used in their work and may object. If a security incident affects a client’s data, we notify that client within 72 hours of confirming it, or sooner where the client’s agreement requires.
Questions
Clients and prospective clients can request our AI Use Disclosure Schedule and supporting assurance materials from our governance contact. Email info@thinkcfsi.com to learn more.
Questions about our AI use are answered from approved language. Anything not covered by this statement or by a client’s AI Use Disclosure Schedule is routed to the President before a response is given, so that clients receive consistent and accurate answers rather than improvised ones.
This statement describes CFSI’s practices as of the date below and is reviewed at least annually. It is provided for information and does not create contractual obligations; client commitments are set out in each client agreement and its AI Use Disclosure Schedule.
