Ask Infrrd #1: Effective OCR Partner for Mortgage

 

Recently, Infrrd launched Infrrd for Mortgage, an AI-led intelligent document processing (IDP) solution specifically designed for the mortgage industry. Our announcement generated a lot of interest from mortgage providers of all sizes and led to a number of interesting conversations. Through these conversations, we discovered quite a few repetitive questions. These seem to be the common concerns among mortgage providers who have not had a good experience with using OCR for mortgage documents or have felt let down with the accuracy of automated processing.

In this blog series, we answer these common questions and make the case for Intelligent Document Processing (IDP) solutions.

Let us begin with the first question.

Q: In the mortgage industry, what are the pros and cons of working with outsourced vendor partners to do data processing vs. investing in the automation process with technology solutions?

A: This is the question that we get asked the most. Many mortgage providers today use a vendor that employs a team of data entry operators who read and enter data from mortgage documents into a Loan Origination Software (LOS).

Using these vendor teams can result in the following challenges:

1. Scaling up for higher volumes - When your business goes through a spike in volume, it is generally difficult for these vendors to find people and train them quickly to keep up with the demand. There is a significant opportunity cost associated with scaling these teams.
   
   
2. Scaling down for slower times - When mortgage rates go up and you see a dip in your processing volumes, you are left with surplus talent that you need to Lay off. We are going through such a time right now. It is much easier to turn off servers than to take someone’s job away and that is another big motivation to use IDP.
   
   
3. Knowledge Churn - Data entry is not a long-term dream job for many people. As a result, there is a lot of churn in these jobs. When a data entry operator walks out of these operations, the knowledge that he/she has accumulated walks out the door. Replacing and retraining new operators drains business dollars as well.
   
   
4. Stagnant Efficiency - There is a relatively low limit on how much efficiency you can get out of a person working for 8 hours. Most data processing operators reach this efficiency within 3 to 6 months. Thereafter, you need to invest a considerable amount in monitoring tools, analytics and managers to make sure that the efficiency does not go down from that high watermark. Technology, on the other hand, keeps improving continuously. If the processing technology is powered with AI and machine learning, efficiency knows no bounds.
   
   
5. Cost - Last but not least, having 100% manual data entry teams is expensive. Depending on the kind of data you process, it may not be legal to send that data out of the country of origin. In this case, the cost increases further by needing to employ a local workforce. Investing in automation technology that is not only compliant but also efficient means long-term cost savings.

Another important reason to consider automation is that your customers will expect you to reduce your processing costs and times significantly just to keep their business. History is full of precedents of automation and how it changed expectations from customers.

Think about this - would you open an account with a bank that does not provide online banking? Or use a cab service that expects you to book it a day in advance? What once started as a differentiator is now a normal service that every customer expects.

AI-based automated, faster, cheaper, and better document processing is a differentiator for mortgage companies today but it will soon become a commodity. You do not want to miss that bus and lose customers by not keeping up.

Here is a lesson from history - This is what the Blockbuster CEO Jim Keyes said about competition from Netflix in 2010:

Today we know how Netflix redefined entertainment and made online streaming the new TV.

We will continue this conversation in our next post where we will talk about How to find an effective OCR partner who can show results.

Do you have a question that you would like us to answer? Please shoot an email to hello@infrrd.ai.