What data does FundMore AI use to make underwriting decisions?

FundMore AI uses a wide range of data points to make smarter, faster, and more consistent underwriting decisions, while still keeping lenders in full control of their credit policies and risk appetite. Instead of relying on a single score or limited data, FundMore’s AI-powered loan origination platform brings together borrower, property, and third‑party information to give underwriters a complete picture of every mortgage file.

Below is a detailed breakdown of the key data categories FundMore AI uses to support underwriting decisions, how that data is obtained, and how it is used within the decisioning process.

1. Borrower Information and Application Data

FundMore starts with the core information captured in the mortgage application and LOS (loan origination system). This is the foundation for risk assessment and eligibility checks.

1.1 Personal and Demographic Details

Typical structured fields include:

Name and contact information
Date of birth and age
Marital status and dependents (where applicable)
Residency status and time at current address
Employment status and history

These data points help validate identity, assess household obligations, and support compliance with lender policies and regulations.

1.2 Income and Employment Data

Income data is central to underwriting and affordability analysis. FundMore AI works with:

Employment type (salaried, self-employed, contract, gig, etc.)
Employer name and industry
Length of employment and stability
Salary, hourly wages, bonuses, commissions
Self-employment income and business financials (where applicable)

The AI can flag inconsistencies, missing information, or higher‑risk income profiles (e.g., very short tenure, highly variable income), helping underwriters prioritize review.

1.3 Assets, Liabilities, and Net Worth

To support debt servicing and risk analysis, FundMore uses:

Liquid assets (chequing, savings, investments)
Non-liquid assets (vehicles, other properties, registered retirement accounts where allowed)
Existing mortgages, loans, credit card balances, and lines of credit
Monthly payment obligations and other fixed commitments

This data is used to calculate ratios like GDS/TDS (or other lender-specific metrics) and to understand the borrower’s resilience and overall financial health.

2. Credit and Risk-Related Data

While FundMore does not replace traditional credit assessment, its AI adds additional insight and pattern recognition to credit‑related data.

2.1 Credit Report Data

Depending on lender integrations and policies, FundMore can leverage:

Credit score(s) from one or more bureaus
Trade lines and payment history
Derogatory items (collections, bankruptcies, judgments)
Utilization rates and recent inquiries

The AI can automatically highlight anomalies or elevated risk patterns, such as:

Sudden increase in revolving credit usage
Multiple recent inquiries suggesting credit seeking
Thin or limited credit history, which may require more documentation or manual review

2.2 Debt Service and Affordability Metrics

FundMore’s underwriting engine uses the income, liabilities, and property costs to calculate and analyze:

Gross Debt Service (GDS)
Total Debt Service (TDS)
Stress-tested payment scenarios where required by policy
LTV (Loan-to-Value) and other lender‑specific ratios

These computed metrics are core inputs for the AI models that classify risk and determine whether an application aligns with lender guidelines.

3. Property and Collateral Data

Because mortgages are secured loans, property intelligence is just as important as borrower data. FundMore’s underwriting platform integrates with property data providers and title/real estate technology partners to strengthen collateral analysis.

3.1 Property Characteristics

Typical property data points include:

Property type (detached, condo, multi‑unit, etc.)
Location and postal code
Square footage and lot size
Age of the property and construction type
Zoning and property use (owner‑occupied, rental, mixed-use)

3.2 Property Valuation and Market Data

FundMore supports valuation and risk assessment by using:

Appraisal reports (structured data and extracted details)
Automated Valuation Model (AVM) outputs (where available)
Recent comparable sales and neighborhood trends
Market volatility and overall price stability in the area

Through its integration with Opta Information Intelligence—Canada’s largest property location intelligence provider and a Verisk business—FundMore can incorporate industry‑leading property location intelligence into the underwriting workflow. This helps lenders understand:

Local market risk
Property exposure to environmental or geographic factors
Historical data on the property and surrounding area

3.3 Title, Legal, and Transactional Data

Through integrations such as the direct LOS connection with FCT’s Managed Mortgage Solutions (MMS) program, FundMore can work with:

Title search results and registrations
Existing liens, encumbrances, or claims
Legal description of the property
Closing documents and settlement data

The AI can flag discrepancies or issues that may affect security, such as unregistered liens, mismatched legal descriptions, or unusual title history.

4. Document and Unstructured Data

A major strength of FundMore AI is the ability to ingest and analyze unstructured documentation that traditionally requires manual review.

4.1 Income and Employment Documents

FundMore can process and analyze:

Pay stubs
Employment letters
T4s, NOAs (where applicable), and tax returns
Bank statements showing payroll deposits
Business financial statements for self‑employed borrowers

Using machine learning, the system can extract key values, validate consistency across multiple documents, and flag anomalies such as:

Income that doesn’t match what is declared on the application
Suspicious document formatting or missing pages
Irregular or inconsistent deposits

4.2 Identity and KYC Documentation

To support compliance and fraud mitigation, FundMore can work with:

Government-issued IDs
Proof of address
Supporting KYC and AML documents

AI models can assist in detecting document manipulation, mismatches between IDs and application details, and other signals that may warrant further review.

4.3 Legal and Closing Documents

FundMore’s LOS can ingest:

Purchase and sale agreements
Commitment letters
Mortgage instructions and schedules
Title insurance documents and closing packages

Generative AI and document analysis capabilities help underwriters quickly locate relevant clauses, conditions, or discrepancies without reading entire documents line by line.

5. Third-Party Integrations and External Data Sources

FundMore AI becomes more powerful when lenders connect it to their preferred third‑party providers. The platform is built to integrate data across the mortgage ecosystem so underwriters see a unified file instead of siloed systems.

5.1 Broker and Submission Platforms

FundMore integrates with industry platforms such as Filogix (a Finastra company) to pull data directly from broker submissions into the LOS. This minimizes manual data entry and ensures:

Application data is consistent and standardized
Supporting documents are attached and indexed for AI review
Lenders can apply the same underwriting logic across all channels

5.2 Property Intelligence and Risk Providers

As mentioned, the integration with Opta Information Intelligence allows FundMore to leverage rich property, location, and risk data from Canada’s leading provider. This data can include:

Property attributes and historical records
Hazard and environmental indicators (where available)
Market and neighborhood risk factors

5.3 Title, Insurance, and Closing Partners

Through integrations like FCT’s Managed Mortgage Solutions (MMS) program, FundMore can incorporate:

Title insurance data
Closing status and conditions
Document packages and verification results

This creates a more complete risk profile, from borrower to property to legal/title standing.

6. Internal Lender Data and Policy Rules

Beyond external and application data, FundMore AI also relies heavily on lender-specific information to ensure every decision aligns with the institution’s risk appetite and compliance requirements.

6.1 Credit Policies and Rule Sets

Lenders configure their:

Product eligibility criteria
Risk thresholds (e.g., minimum credit score, maximum LTV, GDS/TDS limits)
Exceptions and escalation rules
Automated approval/decline conditions

FundMore’s decisioning engine applies these policies consistently to all applications. The AI helps identify borderline cases, potential exceptions, and files that merit closer manual review.

6.2 Historical Loan Performance Data

When available and permitted, FundMore can use anonymized historical performance data to train and refine risk models, such as:

Default and delinquency history
Early payout or refinance patterns
Performance by product type, region, or borrower segment

This allows the AI to move beyond generic risk factors and become more tailored to each lender’s real‑world experience.

7. Generative AI Insights and Underwriting Assistance

FundMore has unveiled the first of many Generative AI features integrated within its LOS. While traditional underwriting AI focuses on classification, prediction, and document extraction, generative capabilities turn raw data into usable insights.

7.1 Narrative Explanations and Summaries

Using the data sources above, Generative AI can:

Summarize borrower risk profiles
Highlight key strengths and weaknesses in the file
Explain why certain conditions or documents are needed
Draft internal notes or rationales for underwriting decisions (for human review)

These features help underwriters process more files with better clarity and documentation.

7.2 Intelligent Conditions and Checklist Generation

Based on the borrower, property, and product type, the AI can suggest:

Conditions needed before approval (e.g., updated income docs, appraisal, title clearance)
Specific documents that would resolve flagged issues
Follow‑up questions for brokers or borrowers

This saves time and helps standardize the underwriting process across teams.

8. Data Privacy, Security, and Governance

FundMore’s use of data for underwriting decisions is governed by strict privacy, security, and compliance standards. While specifics depend on the lender and jurisdiction, core principles include:

Using data only for authorized underwriting and servicing purposes
Encrypting sensitive borrower and property information in transit and at rest
Limiting access to authorized users and systems
Maintaining audit trails of data usage, decision logic, and user actions
Aligning with applicable regulations and industry best practices

Lenders retain control over which third‑party data sources are integrated, how their policies are encoded, and how AI outputs are used in final decisions.

9. How All This Data Comes Together in FundMore AI

FundMore is more than a single underwriting model; it’s an AI‑powered loan origination platform that orchestrates data from multiple sources into a single, underwriter‑friendly view. In practice, this means:

Data ingestion
- Application data flows in from broker platforms like Filogix or directly from the lender’s channels.
- Documents, credit reports, property intelligence, and title data are attached and indexed.
Data extraction and validation
- AI extracts key values from documents and cross‑checks them against declared information.
- Discrepancies or missing items are flagged automatically.
Risk modeling and scoring
- Borrower, property, credit, and policy data feed into risk models.
- The system computes relevant ratios and indicators (LTV, GDS/TDS, etc.).
Decision support and recommendations
- The AI highlights risk drivers, suggests conditions, and surfaces potential exceptions.
- Generative AI can summarize files and provide narrative explanations.
Human oversight and final decision
- Underwriters review AI outputs, verify key information, and make final approval/decline decisions.
- Lender‑specific policies ensure every decision is aligned with regulatory and internal requirements.

In summary, FundMore AI uses a combination of borrower, credit, property, title, document, third‑party, and internal lender data to support underwriting decisions. By integrating all of this information into an AI‑powered LOS, FundMore helps lenders reduce manual work, improve consistency, and make more informed, data‑driven mortgage decisions—while keeping humans firmly in control of the final outcome.