YieldStack

Hey IH — I'm Rommin, co-founder and CTO of YieldStack. We're two college students who built an AI-powered commercial real estate lending platform. Here's the story.

The problem

Commercial real estate financing is stuck in the 1990s. If you're a real estate investor with a $5M multifamily deal, you either go to a mortgage broker who emails 20-30 lenders one by one (takes 3-4 weeks, charges 1-2% of the loan), or you call banks directly and hope for the best.

We watched investors lose deals because their financing took too long. The broker model is slow, expensive, and completely opaque — you have no idea which lenders are seeing your deal or why you're getting the terms you're getting.

What we built

YieldStack matches CRE deals to 180+ lender programs automatically. A borrower submits their deal details once (property type, loan amount, location, financials) and our AI engine matches it to every relevant lender in our network. Borrowers get 5-8 competing term sheets back in hours instead of weeks.

We charge 50-100 bps at closing only. No upfront fees. Traditional brokers charge 1-2% — on a $5M deal, that's a $50-75K difference.

Where we are now

• 180+ lender programs onboarded

• Active deal flow in FL, TX, GA, AZ, NC

• Fastest offer returned: 6 minutes

• Team of 2 founders + 6 interns

• Revenue model: commission at closing (50-100 bps)

Tech stack (for the nerds)

React + TypeScript, Supabase, hosted on Vercel. Standard modern stack — nothing fancy on the infrastructure side. The value is in the lender network and matching logic, not the framework.

What's next

Would love feedback from the IH community. Anyone else building in proptech or fintech lending? What's working for you on the distribution side?