Confer provides an AI-native loan origination system that replaces legacy platforms like Encompass for U.S. credit unions, community banks, and independent mortgage banks. Its suite of nine specialized AI agents automates sales, processing, underwriting, closing, and compliance tasks, allowing lenders to review and approve work while cutting the typical 45‑day cycle to about four days. A 60‑day pilot can demonstrate per‑loan savings of up to $1,700 based on Freddie Mac’s cost‑to‑originate study.
Funding
Funding not disclosed
Founders
Product
Problem
U.S. credit unions, community banks, and independent mortgage banks rely on legacy loan origination systems that require extensive manual effort, leading to long processing cycles of around 45 days and high per‑loan costs.
Solution
Confer offers an AI‑native mortgage loan origination platform that replaces traditional LOS solutions such as Encompass. The system deploys nine specialized AI agents that automate tasks across sales, document processing, underwriting, closing, and compliance while human staff review and approve each step. By handling document classification, data extraction, risk assessment, and compliance checks, the platform reduces the end‑to‑end loan cycle to roughly four days. The AI agents operate in production with auditable outputs, and the platform integrates directly with existing Encompass fields and other lender systems via APIs. A 60‑day pilot lets lenders measure actual per‑loan savings, which can reach up to $1,700 based on Freddie Mac’s cost‑to‑originate benchmarks.
Target Audience
Primary customers are credit unions, community banks, and independent mortgage banks that need faster, lower‑cost loan processing while maintaining regulatory compliance.
Features
- Nine AI agents covering the full mortgage lifecycle: sales, processing, underwriting, closing, and compliance
- Automated document classification and field‑level data extraction for 30‑50 loan documents, averaging 22 seconds per document
- Real‑time risk assessment with credit, capacity, and collateral analysis, including DTI, LTV, AVM reconciliation, and flood checks
- Integrated RAG‑powered chatbot for borrower‑facing interactions grounded in underwriting guidelines
- Open‑source Model Context Protocol (MCP) server for standardized mortgage document processing
- Usage‑based pricing model tied to funded loan volume, eliminating per‑seat licensing
- Full compliance support (TRID timers, HMDA auto‑population, regulatory rule enforcement) with 99.9% accuracy