Cut Your Initial UW Time
Automating Loan Submission through to Initial Underwrite
Abstract
The mortgage industry has long struggled with inefficiencies in the Loan Submission and Initial Underwriting process, leading to delays that extend processing times, increase costs, reduce productivity, and negatively impact customer service. Traditional responses by lenders—such as adding human resources to manage fluctuating volumes, improve quality, and shorten processing timelines—have proven inadequate and unsustainable.
Recognizing these challenges, forward-thinking lenders are investing in intelligent automation to streamline workflows, reduce reliance on manual labor, and build resilient operational frameworks that perform consistently across economic cycles.
Cut Your Initial Underwrite Time from 48 Hours to Under 1 Hour
In our latest blog post, discover how MOZAIQ’s Integrated-AI Mortgage Automation platform addresses these challenges—enhancing accuracy, transparency, and speed in the loan origination process, all without burdening lenders with unnecessary costs.
By automating document indexing, data extraction, audit, and real-time communication between stakeholders via optimized Gen AI-powered solutions, MOZAIQ significantly reduces the time to an initial underwriting decision—potentially from 48 hours to under one hour. The MOZAIQ platform’s modular features empowers loan officers with immediate feedback, supports underwriters by minimizing repeated touches, and provides borrowers with a smoother, faster experience, paving the way toward full-scale digital mortgage transformation.
Continue to read the full blog below, and when you’re done, contact us to learn how MOZAIQ can streamline your mortgage fulfillment operations with Integrated-AI solutions.
Blog Post: Cut Your Initial UW Time from 48 Hours to Under 1 Hour
When a Loan Officer submits a loan package for an Initial Underwriting decision to a Lender (either wholesale or retail), they want that decision to be made as fast as possible, and they want their customer, the borrower, to be happy with the service. Not only does that benefit the loan officer, allowing her to increment her submitted loan count, but also the borrower, who benefits from a faster decision to close on a loan. And what about the lender? It benefits them as well to move the loan through the system as quickly as possible, enabling them to pocket the fees faster, and to minimize the use of their warehouse line. Fast is better.
But there’s always a problem. The loan file is seldom complete, or it’s not accurate, the loan officer not providing all the necessary documents for the borrower type (e.g., self-employed vs. W-2, or multiple bank statements), or simply omitting one or more documents. And the loan officer doesn’t know this until the lender reaches out, sometimes hours, even days, after the initial loan submission. Not good for the borrower, the loan officer, nor the lender.
And, for the lender, not receiving a complete file tends to force the underwriter, the most expensive resource on the payroll, to touch the loan file multiple times, just to issue an initial approval.
Longer lead times, miscommunication, expensive resources, lower throughput. That’s inefficient. That’s cost-prohibitive. But unfortunately this is how the mortgage industry has been operating for decades. Sure, some attempts have been made to automate, introduce technology in search of efficiencies. But there never seems to be enough time, the cyclicality of the mortgage market forcing lenders to put on the brakes in down times, and to forget about long-term solutions when the times are good. Just hire more people. Pay them above-market rates, because they can’t train new ones fast enough, not in a rocketing market. And then the bottom falls out and everyone’s fired, and the lenders have negative loan origination margins, and some even shut down lines of business. Or even their doors.
Automation, with Integrated-AI, are Critical Success Factors
Meanwhile, the smart lenders are not waiting; they are investing in automation now. They’ve learned from their past mistakes—relying on human capital to scale, for one—and so they’re willing to invest in technology and reap the benefits later, when the market will inevitably turn for the better—they’re creating a platform and a business structure that can weather the market shocks, up and down.
That’s where MOZAIQ can help.
Let’s take the process from the Loan Submission (also called Loan Setup or File Review) through the Initial Underwriting Decision to illustrate the benefits of MOZAIQ’s automation solutions.
Loan Submission is the entry point for a loan file into the lender’s system of record. A loan officer compiles the preliminary set of documents from the borrower, validates, to the best of their knowledge, that all the documents are included in the file and that they are accurate, and submits the loan file to the lender through a front-office point of sale (POS) system.
The reality is that the submitted loan package is typically in various stages of completeness, and is seldom ready for an immediate Initial Underwriting decision. Instead, the lender, to ensure the underwriter’s time is not wasted, usually has multiple file review teams—comprised of junior resources—clean up the package, perform data-entry, essentially pre-processing the file before it gets to the underwriter. This inevitably adds time, increases loan origination cost, and slows down the entire origination process.
And, once the documents have been submitted to the lender, the loan officer now is blind, with little visibility into how the loan is progressing through the lender’s system. And if the loan officer is impatient, which they usually are, they’ll pick up the phone and call the lender directly, sometimes calling an underwriter directly, someone who they have a relationship with, asking them about the status of the loan that they had just submitted and by the way can you expedite it through the process?
In general, the entire Loan Submission through Initial Underwrite process for a medium-sized lender takes up to forty-eight hours. Why so long? The initial approval typically results in multiple conditions being generated. And these conditions, instead of being cleared in one session, are piece-mealed to the underwriter, causing them to review a single loan file up to four times before final approval/clear to close.
But what if there was a way to automate critical portions of the process and the communication channels creating efficiencies for all parties involved, compressing timeframes?
Talk to the Loan File
MOZAIQ has expanded its market-leading automation platform to now enable actors—loan officers and lender resources—to interact directly with a loan file. For real. Here’s how it works.
When the loan file is submitted by the loan officer to the lender, it is first processed through the MOZAIQ platform (which is seamlessly integrated with the lender’s Loan Origination System—LOS), and specifically through the Indexing and Data Extraction (IDE) engine. The IDE determines whether all the documents are present in the loan file, and if any documents are missing (missing documents are identified based on the specific borrower and loan scenario identified in the application, and not just the basic “2 months” of paystubs). If there are missing documents in the loan file, the IDE automatically calls the MOZAIQGuard feature, that informs the loan officer of the specific documents that are missing. In real time. No delays. This notification is configurable—email, text, phone call—whatever makes sense for the lender, and the loan officer.
Once all documents are received, the loan file then passes through the MOZAIQAudit feature, that analyses the documents to ensure that there are no discrepancies in the loan file and that the borrower meets the lender’s pre-underwriting criteria. MOZAIQAudit checks the following loan specifications for the Loan Submission process:
Loan Details
Documents
Application
Initial 1003
Income
Paystub, W2, VOE
Assets
Bank Statements, IRA, Retirement Accounts, etc.
Credit Report
Credit Report
Purchase Contract
Purchase Contract, Addendums, Counter Offers
AUS
LP and/or DU
Title
Preliminary Title, Final Title Report
Insurance
Insurance Dec Page
Wire Instructions
Wire Instructions
MOZAIQAudit automatically updates and completes the data in the LOS based on the data extracted from the supporting loan documents. It also ensures data integrity across multiple documents and confirms that lender-required data is available on the supporting documents. The loan submission, including its supporting documentation, is reviewed for the borrower’s Capacity to repay, Capital availability, and Credit Worthiness. The title, insurance, and wire instructions are also validated, and the details are updated to the system of record.
If there are discrepancies in the data, or if business rules flag an issue in the loan submission, conditions, and tasks, are generated by MOZAIQAudit. These conditions, and/or exceptions, can then either be managed by the MOZAIQMonitor feature, or, if the issue is beyond the platform’s capacity to resolve, the loan is kicked out to the lender’s file setup team, and the loan is processed manually through to the Initial Approval.
Meanwhile, as the loan is traversing the system, the loan officer may want to get an update on the status of the loan. They can do so using the MOZAIQAnswers feature, a chatbot tailored by use case, and by user, where they can inquire as to the status of the loan. Likewise, MOZAIQAnswers can be used by the lender operators and underwriters, where they too can: (i) inquire as to the status of a loan; (ii) ask what conditions and/or tasks are pending; (iii) determine what documents are missing; (iv) inquire about specific loan details e.g., is there anything on the borrower’s credit report that might require explanation (from the borrower); (v) or execute a query across an entire pipeline of loans e.g., show all the loans with an incorrect MIP amount?
And, for each process milestone that the loan reaches, the MOZAIQInsights feature is triggered. MOZAIQInsights are pre-configured queries that inform the relevant users as to the status of the loan, proactively providing loan officers, lender operators, and underwriters, with an auditable and trackable loan status as the loan progresses through the origination process.
Here’s a summary of the MOZAIQ features available on the platform, as configured for the Loan Submission through the Initial Underwriting Decision process.
Feature
Description
MOZAIQGuard
- Monitors major milestone changes and automatically escalates unresolved Domain Errors
- For the Loan Setup use case, at the loan stage: ‘Submitted for Initial Review’, a Sentry will run to identify any missing required documents and notify the loan officer of the missing documents and specific data.
MOZAIQMonitor
- Automatically Create, Modify, and clear conditions or tasks as documents and data flow into the system of record (LOS).
MOZAIQInsights
- Domain focused summary of the current state of the loan. For the Loan Setup Case, narratives include the UW Loan Summary and the Occupancy Summary.
MOZAIQAnswers
- Natural language interface to the Audit Domain Services, enabling the Loan Officer, Operations Resource, or Underwriter to query the status of the loan file.
MOZAIQAudit (Loan Submission and Initial UW Decision)
- Provides “Real Context” about every aspect of the loan (updated in real-time)
- Integrated with key services like Freddie Mac’s digital underwriting platform
- For the Loan Setup and Initial UW use case, the following Domain Services are executed:
- Income
- Assets
- Credit Report
- Purchase Contract
- Title (if the document is available)
- Insurance (if the document is available)
- Closing Protection Letter (CPL)
- AUS
- Wire Instructions
- Fraud Reports
And here’s a diagram that shows the MOZAIQ features within the context of the Loan Setup and Initial Underwriting use case.
The more complete and accurate a file is when it is submitted to underwriting, the faster the file will go through underwriting (i.e., credit decision) and on to closing and funding.
Of course the loan type will determine how quickly the loan flows through the system. But for the most common loan products, the goal is for the Initial UW decision to be made within an hour (if not within minutes) of the loan officer submitting the loan file. And, with automation, the feedback loop now in place will continually be used to improve the process, as lenders move toward full automation. This way loan officers, over time, will learn the benefits of submitting a clean loan package the first time.
For more complicated loan products or income scenarios—e.g., non-QM—a lender may only require a junior resource to facilitate the automation in making an initial UW decision. Some loans will require exceptions for indexing and data extraction, that’s inevitable (e.g., for poorly scanned documents), or for loans where borrowers are self-employed or have complicated income/asset structures.
Where To Start
MOZAIQ is working with at least four lenders to implement the first part of this process, the Loan Submission. And they’re doing it the right way, starting with a subset of loan documents, what we’re calling Loan Submission Lite, to kickstart the process, rapidly deploy the automation, see it in action, optimize the workflow, and then add the rest of the documents to the process, before the Initial Underwriting Process is implemented. That way the customer has the baseline and foundations setup correctly, they’ve introduced automation in a manageable fashion, mitigating change management risks (and in mortgage there are many), and ensuring that they are ready for when the market volumes inevitably will increase.
Here is an example of some of the documents these customers are starting with:
- Purchase Contract
- Insurance
- Title
- Wire Instructions
- A light overview of the Assets, Income, and Credit i.e., not full calculations
The key is that lenders can choose where to start, they pick a subset of the documents, and then incrementally add the full lot as they become comfortable with the process.
MOZAIQ Enables Integrated-AI Mortgage Automation
How beneficial is it to a loan officer, a lender, and a borrower to go from an average initial loan approval time of 48 hours to less than one hour? How valuable is it for a lender to enable scale with automation whether it’s a falling or rising mortgage market? How critical is it for lenders to provide superior customer service to their loan officers and borrowers? How important is it for a lender to see efficiencies translate directly to their bottom line?
With MOZAIQ’s Integrated-AI Mortgage Automation solutions, give loan officers the tools they need to interact with the submitted loan and to query its status, anytime, anywhere. Enable the lender to process the loan file faster, with minimal human intervention, and maximize the efficiency of the underwriters—the most valuable resource on the team. And, give the borrower peace of mind, knowing that their documents have been accepted, and their loan file is complete.
If you’re ready to achieve real business benefits with a winning automation strategy, follow the lead of Mortgage Lenders who choose MOZAIQ. Contact us today and discover how our Integrated-AI, End-to-End, Intelligent Mortgage Automation solutions can help you win.
Note: This blog post was written by a human and does not contain content generated by ChatGPT or any other Generative-AI platform.
