The Situation
A leading Australian online mortgage broker had an blockage in their conversion pipeline: the time it took to verify eligible applicants. Before working with Remi AI, the application process was as such:
-
Applicant uploads relevant documents including drivers license, payslip, passport, bank statement, utility bills.
-
Verifier receives documents and files said documents to relevant location
-
Verifier manually extracts information from each documents.
-
Verifier enters data into a master database
This was a laborious data entry process for the ~ 6000 applications per month. The task averaged 60min of processing time per applicant.
The Approach
Remi AI was engaged to speed up this process, and we did this through the use of Intelligent Document Extraction (IDE).
Once the IDE process was successfully added to the existing pipeline, the application verification process was reduced to a little over three minutes - namely for the human verification of the date input. This is a time saving of 2000%
What is Intelligent Document Extraction?
Intelligent Document Extraction (IDE) is a software solution purpose built for businesses needing to extract specific pieces of information from a variety of documents including those that have been scanned to PDF. IDE does this through the following steps:

Existing Pipeline
Where IDE steps in
Applicant uploads requisite documents
Verifier sorts documents
Verifier manually extracts relevant data
Verifier enters relevant data into system
Successful applicant moves down pipeline

iDE
The Results
Remi AI was able to reduce the average hour-long applicant process down to 3 minutes. This is an improvement of 2000%
We were able to utilise our clients existing pipeline, so there was minimal need for retraining or additional software costs.
Is IDE for you?
Remi AI's IDE tool is the perfect solution for application processes, screening, on-boarding, or any data extraction process where there is a uniformity to the to documents being use. If you have custom documents, ie, a business specific application process, IDE can easily be trained on any number of new documents.