HousingWire: How Tidalwave Helped First Colony Mortgage Scale Wholesale to $600M Without Adding Headcount

First Colony Mortgage is the first lender to run both wholesale and retail on Tidalwave's agentic AI platform. Here's what changed in their workflow, and the year of production work that got them there.
April 29, 2026

The U.S. mortgage industry has had a decade of "digital transformation" and not much to show for it. Loan files still routinely exceed 500 pages. Approvals still take 43 to 45 days. The average retail lender lost $645 per loan in Q1 2024. Most of the work that drives those numbers is still humans copy-pasting between fragmented systems.

Wholesale lending is the ugliest version of that problem. A wholesale lender doesn't originate the loan directly. They review packages submitted by outside brokers, and those packages can run 500 to 1,000 pages: bank statements, pay stubs, tax forms, a 1003, the whole file. Manually reviewing all of it is one of the most time-intensive parts of the business.

This is the part of the mortgage workflow First Colony Mortgage decided to rebuild.

Why First Colony brought in Tidalwave

First Colony came to Tidalwave about a year ago with a specific problem: their wholesale division was growing faster than they could staff it. Margins were thin. Hiring more underwriters wasn't a viable answer.

In a recent interview with HousingWire at The Gathering, First Colony's Corey Shelley described the moment:

"With wholesale, we were growing so fast, and margins are thin and scalability is hard. What we saw with Diane's product is a way that we could scale, cut out a lot of time, without having to keep hiring and hiring more people."

Most "AI in mortgage" stories stop at the borrower-facing portal. First Colony was looking for something else: production AI that could touch the back-office work driving their cost structure.

What we built together

Tidalwave and First Colony spent a year building two capabilities specifically against First Colony's production volume:

Doc assist reads the entire loan package the moment it lands, categorizes every page, and flags what's missing or inconsistent before it ever reaches an underwriter.

Asset assist handles forensic bank statement analysis. That analysis typically takes an underwriter two to four hours per borrower. On Tidalwave, it completes in five to six minutes.

Both features have since rolled out to all of Tidalwave's point-of-sale customers.

The work wasn't done in a sandbox. As Tidalwave CEO Diane Yu put it in the same interview: "There's a huge difference between a demo and the real production product. A production product has to take all those complex scenarios to be laser accurate."

What changed in the workflow

Before Tidalwave, First Colony's wholesale team spent the early part of every file on what Corey called "stare and compare": manually organizing 500 to 1,000-page packages and looking for inconsistencies across documents.

Now, the agent does that work the moment the package arrives. It categorizes every page, correlates data across credit, income, and bank statements, and surfaces issues before a human opens the file.

Corey's framing in the interview captures the shift:

"You kind of have to break how you think about mortgages. Underwriting starts the second the borrower uploads everything."

That's not a faster mortgage. It's a different operating model.

The results

Eighteen months in:

• First Colony scaled wholesale volume from near-zero to $600 million
• 100% of First Colony's wholesale volume now runs on Tidalwave
• Forensic bank statement analysis: two to four hours, down to five to six minutes
• First Colony is now the first Tidalwave customer running both wholesale and retail on the platform

The change isn't just operational. It's showing up in how the team talks about the work. Corey on the response from his own people:

"My loan officers love it. I'm getting messages and texts constantly saying, 'Hey, this is the greatest thing that we've ever used.' My underwriter is saying we're saving so much time. The AI is showing us everything we need to see immediately."

Loan officers and underwriters are the two highest-cost roles in a mortgage lender's P&L. When both are reporting time savings unprompted, the system is working.

What this points to

The First Colony deployment is one of the clearest production examples of what agentic AI can do inside a regulated workflow. Not a demo. Not a borrower-facing chatbot. The actual back-office work that drives cost, margin, and turn time.

It also points to where the industry is going. Most lenders today are still running the wholesale process as a series of manual reviews stitched together by email and spreadsheets. The technology to do it differently exists. First Colony is one of the first to commit to it at scale.

For the lenders watching: the question isn't whether agentic AI will reach wholesale and retail mortgage operations. It's how much volume your competitors will run on it before you do.

Lenders, schedule a demo here: https://www.tidalwave.com/book-a-call

Watch the full HousingWire interview with Diane Yu and Corey Shelley: https://www.housingwire.com/videos/from-chaos-to-clean-files-what-happens-when-ai-meets-a-700-page-loan-file/