-
-
Equity Analysis done right!
-
-
-
-
-
-
Inspiration
After thorough market research, we have realized that the Equity Analysis (EA) of any loan is a time-consuming process, and Bank users of the Real Estate department need to manage the process manually. This market analysis inspired us to create a Digitized & Real-time Equity Analysis process with automation and meet the demand that eventually will reduce overall processing time and prevent considerable loss to Banks. Though we have created the use cases of the Solution based on Loan Default applications like Foreclosure, the same Solution will be good fit for the other Loan processes like Loan Servicing, Loss Mitigation, etc.
Research on EA for HELOC : A HELOC (Home Equity Line Of Credit) is similar to a business line of credit. The lender uses the house as security and provides a homeowner with a line of credit that has a fixed limit. The borrower can withdraw funds within a time known as the “draw period” which is on average 10 years.
Housing prices increased in the past few years, and hence HELOC collateral values were
high. Most of the loans are still in the Draw period. Customer paying only interest.
Collateral value changes with the economy and market condition, which means some loans might have a negative or low equity value. But, Financial Institutes & Banks cannot do Collateral Equity Analysis process proactively due to static collateral valuation, a complex, time-consuming, and involved costly collateral appraisal process that takes around 30 days to complete.
What it does
The Collateral Equity Analysis Advisor is a Pega-powered PropTech solution to facilitate real-time Equity Decision of a Loan. It reshapes the traditional service model of Collateral document verifications and Property valuations with disruptive new technologies and innovative ideas that seek to revolutionize the real estate markets. Earlier, the Bank department employees manually reviewed Collateral/Lien documents and relied on Human appraisers to receive the property valuation information. Once valuation details are received, they manually calculate the loan's Equity Decision.
The Application provides a 100% Automated & Augmented process to complete above mentioned process with a reduction of 70% of the total time. The Bank's internal collection users of the Application gather the Borrowers' delinquent loan information and transfer the information to the Foreclosure or Bankruptcy or Loss Mitigation Specialist. The Application creates a Case per loan and assigns it to the Queue. The Specialist pulls the Cases from Queue to review Borrower/Loan information and uploads the collateral/Lien documents in PDF/Image/Doc format. The Application has integration with Azure OCR APIs to get the raw text data from PDF/Image/Doc. An internal Pega NLP module extracts the required information from the raw data to present on user's screens. In this way, the Application reduces significant time to key in Collateral data manually. The screens are designed in such a way that critical data like amounts, addresses are highlighted, and these friendly screens will allow Users to verify the data easily. The Specialist confirms the information and submits the case to get the property's appraisal information like Valuation amount, date, and others. The Application has integration with external valuation APIs like ATTOM, Datafiniti to capture property details. Once API returns the data, Business decision module compares the loan balance with appraisal data and calculate appropriate Equity Analysis of a loan. In this way, the Application eliminates the Human-centric Appraisal process and Equity Analysis.
How we built it
We have had a design thinking session to create business cases and decomposes into user stories and loads them into Agile Workbench. The application has three major roles - Collection, Specialist, Manager. It is divided into separate modules (by Ruleset) to leverage agility. 1. Module 1 : Business Process 2. Module 2 : Document Extraction (Using OCR/NLP) 3. Module 3 : Equity Decision (Using Appraisal APIs)
- Module 1: It consists of Case Management, Screen, and the Business Logic to process cases from one user to another.
- Module 2: The OCR/NLP module process each document. The raw data of the document are extracted by OCR APIs and stored in a table. Pega NLP takes data from the table and performs text analytics using the "entity extraction" technique. The Module keeps the required information in appropriate columns of the table to display on the screen.
- Module 3: This Module exposes REST services to integrate with Business modules. It has a Case Management process too. The Business module passes required Loan/Collateral/Lien information to invoke Appraisal APIs and retrieve property valuation. Then, calculate the complex Equity Analysis logic to decide the loan's Decision and store in a table to display results on the User screen.
Challenges we ran into
When we decided to build a PropTech solution, our objective was to overcome a particular Default business process's complexities and reduce overall processing time through the digital journey. So, we had to spend significant time on Research & Discovery phases.
Additionally, we had several brainstorming sessions to discuss building the OCR component. There were few options available, and we'd decided to go for Azure because of its high accuracy, performance, and simple integration.
Lastly, it took us some time to find appropriate Appraisal APIs to give us the valid data feeds we need, especially Sale Comparable. Here we have integrated with ATTOM/ Datafiniti APIs.
Accomplishments that we're proud of
First of all, we started with very little information making a FinTech type solution. The Team collaboratively worked, been through many design thinking sessions, and came up with innovative ideas, automation Use cases & Business benefits to redefine the ways the property is managed and marketed. We loved that we had explored many emerging technologies like Optical Character Recognition, Natural Language Processing, Property Valuation APIs and so on, and we successfully implemented them in our solution. The orchestration of different modules/components of the application is the backbone, and improved performance is the application's muscle.
What we learned
We've learned how App Studio and Agile Workbench can be an essential part of application development, so the Business and Project Team can work together collaboratively. We worked virtually to develop the application and learned to use Collaboration tools like MS Teams for quick discussion, decision instead of sending emails or documents. Also, the opportunity gave us to know the power of Design Thinking and Pega's Text Analytics features NLP.
What's next for Collateral Equity Analysis Advisor
This PropTech Application is a model application and ready to install in Loan Servicing or Default Ecosystems. IT & Business team can enhance the application based on needs. One of the enhancements that we're analyzing is "Virtual inspection". We'll use Pega mobile so that the customer on its own can capture pictures of the interior and exterior. We can use proximity services like Geocoder API to determine if the customer is at his/her address.
Log in or sign up for Devpost to join the conversation.