Loan Boarding
Objective:
The objective of this project is to integrate Optical Character Recognition (OCR) and Machine Learning (ML) technologies into the application to automate the process of updating information from PDF documents directly into the application fields. This will streamline data entry processes, reduce manual errors, and improve overall efficiency.
Scope:
The scope of this project includes the development and implementation of OCR and ML algorithms within the application framework. Specifically, the system will be able to:
Identify and select relevant information from PDF documents. Eg if we have 500 pages bunch. Only 10 Document need to be identify and scanned, those document name are as follow:
- First Lien Doc (Deed of Trust)
- Second Lien Doc (2nd Deed of Trust)
- Allonge to Note
- Modification of Mortgage
- Equity Line Credit Agreement
- Assignment of Mortgage
- Promissory Note
- Adjustable Rate Note
- Success & Entrent Deed
- Title Doc
- Notice of Foreclose
OCR technology will read the information and Copy the selected information from the PDF.
Select the corresponding field where to Paste the copied information OCR Technology will Paste the information into the corresponding fields within the application.
Implement manual verification by the processor to ensure accuracy, correcting any spelling mistakes or special characters encountered during the process.
Functional Requirements:
OCR Integration:
The system shall be capable of extracting text from PDF documents accurately using OCR technology.
It shall identify and highlight the relevant information within the PDF.
The processor shall have the option to select and confirm the highlighted information for extraction.
ML Integration:
The ML algorithm shall analyze the extracted text to identify the appropriate application fields for data entry.
It shall automate the process of pasting the extracted information into the corresponding fields within the application.
Manual Verification:
The processor shall manually review the extracted information for accuracy.
They shall correct any spelling mistakes or special characters encountered during the extraction process.
Non-Functional Requirements:
Accuracy:
The OCR and ML algorithms shall strive for a high level of accuracy in extracting and updating information.
The manual verification process shall serve as a quality control measure to ensure data accuracy.
Performance:
The system shall perform efficiently, with minimal latency in processing PDF documents and updating application fields.
User Interface:
The user interface shall be intuitive and user-friendly for both processors and administrators.
It shall provide clear instructions and feedback during the OCR and ML integration process.
Assumptions and Constraints:
Assumptions:
The PDF documents provided will be of standard format and layout, facilitating accurate text extraction.
Processors will have the necessary training to perform manual verification effectively.
Documents may vary in format and layout, potentially posing challenges for accurate extraction.
Constraints:
The system's performance may be affected by the quality and clarity of the PDF documents.
Integration with legacy systems or third-party applications may pose compatibility challenges.
Risks and Mitigation:
Risks:
Inaccurate OCR or ML processing leading to data entry errors.
Insufficient manual verification resulting in overlooked mistakes.
Mitigation:
Regular testing and validation of OCR and ML algorithms to improve accuracy.
Implementing a robust manual verification process with adequate training for processors.
Support
Should you need any help with our platform get in touch with us
Knowledgebase
Using any of our products and need help? Get in touch with customer support