SAN JOSE, California, May 8, 2023 – Existing home sales surged to a fifteen-year high in 2021,
an 8.5% increase from 2020, with 6.12 million homes sold. However, primarily due to a rise in
interest rates, only 5.95 million homes were sold in 2022.
Conventional mortgages are residential loans comprising 82% of the mortgage market in the
United States. Despite being the most common type of mortgage, a conventional loan is difficult
to obtain because borrowers need to qualify regarding many factors, which include having a
minimum credit score of around 620, a debt-to-income ratio of no more than 43%, and a down
payment of at least 20%. In addition, their credit reports should not have any significant issues
like bankruptcy. To achieve these goals, borrowers need to provide many documents like W-2s,
pay-stubs, financial statements, tax returns, and verification of employment letters. This makes
the entire process extremely unwieldly, costly, and manually laborious. In fact, a recent report
shows that this cumbersome process, low housing inventory, and affordability challenges in
2022 forced lenders to lose an average of 301 Dollars per loan. Keeping this in view, Scry AI’s
software, Collatio®, which is a leader in Intelligent Document Processing (IDP), provides a fast,
inexpensive, and accurate process for handling conventional loan applications.
Collatio® uses more than forty proprietary AI-based algorithms and pre-built ontologies that
enable the classification of around sixty types of loan-related documents. These documents
include W-2s, bank statements, tax returns, credit reports, preliminary title reports, appraisal
reports, property tax bills, homeowner’s insurance policies, and various financial statements.
Collatio®’s AI powered algorithms classifies all incoming documents, and immediately lets the
borrower know the ones that are missing or are of bad quality. Since this classification and
validation reduces the overall time from several hours to a few seconds, it provides a splendid
experience to the borrowers. It also saves lenders more than 80% of their cost and time spent
asking borrowers to provide the missing documents.
According to Dr. Alok Aggarwal, CEO and Chief Data Scientist of Scry AI, “Since Collatio®’s
advanced AI trained models can classify incoming documents very accurately and very quickly,
it alerts the borrower immediately regarding missing or bad quality documents. Hence, it is a
blessing because borrowers save time during this agonizing process and lenders do not need to
repeatedly request them for missing documents.”
After classification, Collatio® extracts the required key-value pairs from each document and uses
AI-based algorithms to reconcile the extracted values. For instance, it reconciles the total
income from pay stubs, W-2 forms, and tax returns. Because of AI-based reconciliation, it
achieves straight-through processing for more than 60% of the input documents, thereby freeing
up analysts’ time so that they can attend to more complex tasks. Furthermore, it provides alerts
regarding abnormal situations such as errors in the paperwork or potentially fraudulent
“While Collatio® provides valuable insights and data-driven analysis, its human-centric approach
can help mortgage advisors to offer more personalized, timely, and effective services to their
customers. It also helps credit underwriters spend more time on higher value-added tasks,” said
Srinivasan Bharadwaj, Head of Digital Transformation of Scry AI.
Finally, Collatio® can be integrated with almost all end-to-end software lending solutions and
processes and it comes with a configurable set of APIs and a rich user interface. It is sold in
SaaS (software as a service) mode but can also be installed on-premises and behind a firm’s
information technology firewall.
Company Details –
Scry AI, Inc. (“Scry AI”) was founded in 2014 and builds innovative AI-based enterprise
applications that enable clients to rethink and automate their data-driven and manually intensive
business operations. Scry’s families of enterprise apps include Collatio® (for ingesting,
extracting, and reconciling unstructured and structured data), Anomalia® (for detecting
anomalies and potential fraud), Concentio® (for ingesting and harmonizing IoT data), and Vigilo®
(for predicting operating and marketing risks). Further details can be found at www.scryai.com.