🚀 Another important milestone for MortgageMind AI — DTI is now working end-to-end.
Today I completed the next piece of the underwriting workflow in my MortgageMind AI project.
The system can now take verified borrower income and calculate the Debt-to-Income (DTI) ratio directly inside the loan workspace.
The workflow now looks like:
📄 Paystub uploaded
↓
🔎 OCR extracts the document text
↓
🤖 AI analyzes the paystub
↓
💰 Gross income is extracted and verified
↓
📊 Monthly debt obligations are captured
↓
🧮 DTI is calculated automatically
↓
🏦 DTI becomes part of the underwriting workspace
A real example from today’s testing:
Verified monthly gross income: $15,500
Monthly debt obligations: $4,650
DTI = $4,650 ÷ $15,500 × 100 = 30.00%
The important part for me isn’t simply calculating 30%.
It’s connecting the entire chain:
Document → AI → Structured Data → Verification → Financial Calculation → Underwriting
The DTI result is now persisted with the loan and displayed alongside the verified income and source document in the MortgageMind workspace.
This is bringing the project much closer to the type of workflow I want to build for modern mortgage operations — reducing manual data movement while keeping the underlying calculations transparent and auditable.
I’m deliberately keeping the scope focused rather than continuously adding features.
One underwriting workflow at a time. One measurable improvement at a time.
#MortgageMind #MortgageTechnology #MortgageAI #AI #ArtificialIntelligence #MortgageUnderwriting #DTI #DigitalMortgage #FinTech #PropTech #LoanOrigination #MortgageLending #Underwriting #Automation #DocumentAI #GenerativeAI #OCR #Python #FastAPI #NextJS #PostgreSQL #SQLAlchemy #SoftwareEngineering
