Copperlane Puts an AI Loan Officer Inside Every Mortgage Application

The Macro: Mortgage Origination Takes 45 Days and Nobody Knows Why

The average mortgage takes 45 days to close. Forty-five days of document collection, verification, back-and-forth communication, and manual data entry. The borrower submits W-2s, bank statements, tax returns, and pay stubs. The loan officer reviews them, asks for corrections, and enters data into the origination system. The underwriter reviews the file, asks for more documents, and the cycle repeats.

Most of this time is not spent making decisions. It is spent processing paper. A loan officer might spend 60% of their day chasing documents, sending reminder emails, and manually entering information from PDFs into software systems. The actual credit analysis and decision-making takes a fraction of the total timeline.

The mortgage technology market has modernized some parts of this. Blend and Encompass handle the origination workflow. Reggora automates appraisal management. But the intake process, the initial document collection and verification that happens before underwriting even begins, remains largely manual.

The Micro: An AI Agent Named Penny That Handles Intake

Athan Zhang and Brianna Lin cofounded Copperlane. Athan studied CS at Princeton and was previously a quant developer and 2x founding team member. Brianna studied CS and Finance at Penn’s M&T program and was a former Wall Street trader. They are a two-person team from YC Winter 2026 with Harshita Arora. The company operates as Coevolved, Inc. doing business as Copperlane.

The product is an AI-native loan origination system featuring an agent called Penny that automates the borrower intake process. Penny collects documents, asks relevant follow-up questions, auto-fills application forms, extracts data from W-2s and bank statements, pre-screens documents for authenticity, and flags issues before underwriting.

The borrower experience is designed to be simple. Auto-filled forms show only relevant fields. The AI provides instant answers 24/7. Documents get processed as they arrive rather than sitting in a queue. For loan officers, a single-screen pipeline dashboard shows application status, document counts, and flags across their entire portfolio.

The document verification layer is particularly interesting. Pre-screening for authenticity before underwriting catches problems early, reducing the painful cycle of submission, rejection, and resubmission that drags out timelines.

The Verdict

Copperlane is targeting the right bottleneck in mortgage origination. Document intake is where the most time is wasted and where AI can deliver the most value. The founding team has both the technical depth and financial services understanding to build this well.

The competitive risk is significant. Blend is well-funded and already automates parts of the mortgage workflow. ICE Mortgage Technology (formerly Ellie Mae) dominates the origination platform market. If either adds AI intake capabilities comparable to Penny, Copperlane faces tough competition. But both are large companies with complex product lines, and a focused startup can iterate faster on a single workflow.

In 30 days, I want to see the average document collection time. How many days does it take Penny versus a human loan officer? In 60 days, the question is lender adoption. How many mortgage companies are using Copperlane in production? In 90 days, I want to know about the error rate. If Penny catches document issues that human loan officers miss, the underwriting team becomes a champion for the product, and that internal advocacy drives expansion.