About
The mortgage industry remains one of the largest, most antiquated, industries on the planet earth. We could no longer just sit back letting other humans charge you hand over fist for less than subpar services.
That's why we created ArmorDoc! Us humans, along with aliens, teamed up to solve the biggest pain point in the mortgage industry today, documents. We classify, organize and extract data from messy mortgage documents allowing you to focus on originating, purchasing and servicing more and more loans with less and less pain because, well, why not?
So come join us for a pain free journey into our universe to solve this messy Earth issue.
Keep on scrolling down to find out more!
How We Help
-
We take care of all your document related heavy lifting so you can focus on the important stuff
-
We put an emphasis on accuracy, paying close attention to detail
-
We provide transparency such as missing, signed and stamped documents
-
We deliver results fast, at scale
-
We cater to your needs and seamlessly integrate into your workflow with minimal, to zero, disruption
-
We've even built a document management system for you so you can easily find all your documents quickly
-
To sum it up, we are here to help you!
Our Services
We ingest loan packages in bulk, process them through our proprietary document machine learning pipeline which results in the identification, classification and collation of each mortgage document. The result of which is a fully bookmarked PDF file and a missing document report.
Document Recognition
We ingest loan packages in bulk, process them through our proprietary document machine learning pipeline which results in the identification, classification and collation of each mortgage document. The result of which is a fully bookmarked PDF file in a custom stacking order and a missing document report.
Various deep learning and computer vision models have been trained to extract key data points from various documents with a high level of precision. These data points are then compared to each other to determine the final data points and to any external data files provided.
Data Extraction
Computer Vision models are used to confirm the existence of signatures to verify documents have been executed.
Signature & Stamp Recognition
Custom document management system to store, organize and locate your documents in a secure, user-friendly web application.
Document Management System
Industry
Primary and Secondary Residential Mortgage Market
Clients
Banks, Broker/Dealers, Custodians, Hedge Funds, Investment Banks, Lenders/Originators, Portfolio Managers, Servicers
Loan Types
Agency, Jumbos, MSRs, Non-QM, NPLs, Reverse Mortgages, RPLs, Qualified Mortgages
Use Case Examples
Buyside: assist in upfront due diligence to quickly identify document and data issues to reduce risk
Custodians: expediate the collateral review process to quicky identify missing documents
Lenders/Originators: extract data from origination and borrower documents to upload into LOS
Portfolio Managers: assist in the disposition of assets to make more informed hold/sell/securitize strategies
Reverse Mortgages: seamless file preparation for HUD endorsements
Servicers: extract data from borrower documents to seamlessly onboard new loans onto your servicing platform
Our Technology
-
Sophisticated: we combine deep learning and computer vision to build and train models specfic for mortgages
-
Modular: built in modular blocks so individual components can be assembled together into a workflow which meets your business process needs
-
Continuous Learning: assisted by human-in-the-loop feedback, the accuracy of our models are constantly evolving
-
Generalizable: our models can be applied across all loan and document types
-
Integrations: we are able to connect to your system of record to deliver results via API capabilities
-
Document Management: our client portal stores your documents and data in a user-friendly web application
Team
Mike Hartman
CEO and Co-Founder
15+ years of experience in the secondary residential mortgage industry with a focus on data management and due diligence.
Prior work experiences consist of VP of Data, Senior Account Executive and Senior Product Specialist at the likes of JP Morgan, Allonhill and dvo1, respectively.
Brandon Joseph, PhD
CTO and Co-Founder
10+ years as a Data Scientist with a focus on Machine Learning. Ran a large team of Data Scientists at a nationwide health insurer with a focus on predictive analysis and artificial intelligence.
Obtained PhD from U of Maine with emphasis on quantitative methods and former faculty member at U of South Carolina.
Masters in Applied Data Science from the University of Syracuse. Yuwen focuses on building and training machine learning models with a specialization in Natural Language Processing.
Yuwen Hsieh
Data Scientist
Masters in Applied Data Science from the University of Syracuse. Yuwen focuses on building and training machine learning models with a specialization in Natural Language Processing.
Tom Bishop, PhD
Investor/ Advisor
Computer vision, Machine Learning and Data Science expert with 10+ years of industry experience at Apple and several startups.
PhD in signal/image processing, U of Edinburgh. Masters of Engineering. Cambridge.
Guy Shechter, PhD
Chief Architect
Guy has a PhD in Biomedical Engineering from Johns Hopkins University and was previously the Head of Innovation Strategy at Philips.
After 15 years in corporate leadership, he's transitioned to a more hands-on environment focusing on creating value from data, using the latest advances in cloud computing, analytics and machine learning algorithms.
Matt Pitman
Data Scientist
Graduate from Rhode Island with a concentration in Computer and Data Science. Matt's focus is primarily with building data extraction models, however, is always willing to roll up his sleeves to dive into solving any problems.
Former Intern at ArmorDoc
