customer stories
Key Mortgage Services reduces document indexing time by up to 67% with Ocrolus
Customer Stories
Key Mortgage Services reduces document indexing time by up to 67% with Ocrolus
Mortgage | Classify and Analyze
Key Mortgage Services is an independent, family-owned, full-service financial products provider. With a focus on helping buyers save money and close on time, the team relies on the latest technology to keep its operations competitive.
To keep pace with loan volume, the team turned to Ocrolus, a vertical AI workflow and analytics platform for lenders, to automate document classification directly within Encompass. Today, Key Mortgage's setup team reviews only a handful of unclassified documents per file, and Ocrolus handles the rest.
Challenge
Key Mortgage Services relied on manual document indexing and third-party support to prepare loan files, creating a time-intensive bottleneck for its setup team.
Solution
Ocrolus automated document classification within Encompass, giving the setup team organized, ready-to-review loan files from the moment a loan is submitted.
Result
• Up to 67% reduction in classification time: Indexing dropped from 20-30 minutes to 5-10 minutes per file
• Increased loan throughput without added headcount: The setup department handles more files with the same team
• Eliminated manual stare-and-compare verification: Cross-checks that once required staff to manually confirm each VOE across the 1003, AUS and loan file now happen automatically
Challenge: Manual indexing slowed every file
Before Ocrolus, every loan file required a third-party vendor to access and manually sort each document into the correct folder, a process that took 20 to 30 minutes per file.
"Indexing was manual," said Luba Mainz, System Administrator at Key Mortgage. "We used a third party that would go into the file, basically drag and drop things where they go."
The setup team also spent significant time on what Mainz called "stare and compare": manually verifying that a VOE appeared correctly across the 1003, AUS and loan file. The process added friction to every loan and left room for error.
Solution: Automated classification of documents
Key Mortgage deployed Ocrolus to automate document sorting within Encompass. When a loan is submitted to fulfillment, Ocrolus classifies the documents and places them into the correct locations before the setup team begins its review. Staff now work through the unknown folder, typically five documents or fewer, and move straight to ordering services.
The implementation was fast and self-directed. Ocrolus's document mapping feature let Mainz correct classifications in real time without routing changes to a support team. "It was very easy," said Mainz. "Right as we implemented, Ocrolus came out with the document mapping that we can do ourselves, which was way quicker than trying to reach out to people."
Ocrolus also calculates income on every file, giving Key Mortgage a foundation it is actively building on as it works to integrate income analysis more fully into its workflow.
Result: Faster setup, more capacity and no more stare-and-compare
With Ocrolus handling classification, the setup team now processes files in a fraction of the time. "Now it's probably like 5 to 10 minutes versus the half an hour before," said Mainz.
Key Mortgage has increased the number of loans flowing through setup without adding staff. "We've been able to increase how many loans go through our setup department without adding people because it's not a human that's doing the indexing."
The most meaningful change has been the elimination of manual stare-and-compare work. "It's the getting rid of the 'stare and compare,'" said Mainz. Before Ocrolus, staff had to manually verify that each VOE appeared consistently across the 1003, AUS and loan file. Now those cross-checks happen automatically.
Key outcomes:
- Up to 67% reduction in classification time: Indexing dropped from 20-30 minutes to 5-10 minutes per file
- Increased loan throughput without added headcount: The setup department handles more files with the same team
- Eliminated manual stare-and-compare verification: Cross-checks that once required staff to manually confirm each VOE across the 1003, AUS and loan file now happen automatically
Why Key Mortgage recommends Ocrolus
Trust in the platform has grown steadily, particularly around classification. "People can see that it's working and that it's faster than a human doing it alone," said Mainz.
For lenders evaluating AI, Mainz's advice is direct: vet the vendor, then commit to trusting the technology. "You have to vet your vendor, but then you have to trust the vendor and realize that yes, AI can make mistakes, but you need to trust that it's doing what it's supposed to be doing."
Key Mortgage also highlighted the support it has received from day one. Its support team has been "very quick to respond and assist in anything."
Looking ahead, the team is building toward broader use of Ocrolus income analysis — working through how to introduce it into their workflows. "Income's definitely the next step," said Mainz. "We're working through just how we fit it into the workflow and continue to build trust with AI and automation.”
“Before Ocrolus, document classification would take the team about 20 to 30 minutes. Now it takes 5 to 10 minutes versus half an hour before.”
Luba Mainz, System Administrator at Key Mortgage Services, Inc.
