SM
Sheby M
Loading portfolio experience
Real Estate, Finance · AI Agent, Automation, Dashboard
BridgeUW AI – Deal Screening & Underwriting
An AI-assisted underwriting workflow that extracts healthcare and multifamily deal information, applies deterministic Quick Kill rules, calculates available deal scores, and produces an auditable PASS, FAIL, or INCOMPLETE response.
Stack and focus areas
n8n/MakeAIAutomationPythonOpenAIn8nEmail ProcessingOCRPDF ProcessingExcel ProcessingUnderwritingRule EngineDeal ScoringFEMACensus APIRisk AnalysisSource CitationsHuman ReviewAudit Logs
Project description
Healthcare and multifamily opportunities arrived through emails and attachments containing inconsistent information. Underwriters manually searched for sponsor liquidity, background issues, floodplain status, liens, market data, construction protections, and other inputs before applying Quick Kill and scoring rules.
I developed an intake and scoring workflow that extracts required values from email bodies, PDFs, spreadsheets, and images. Every value retains its source, and absent information is marked INCOMPLETE rather than assumed. When authorized external sources are configured, the workflow can retrieve location, flood, population, or public-risk data. Deterministic rules evaluate Healthcare and Multifamily Quick Kills. Passing deals receive a score using available values, followed by a structured summary for BridgeUW and human review.
