Intelligent Data Extraction
ABDP reads your
documents and extracts key information automatically. The system uses a
combination of:
● Text recognition to read printed and
typed text
● AI models to understand context and find
the right information
● Pattern matching to identify dates,
amounts, and IDs
What Gets
Extracted: Depending on your queue configuration, ABDP can extract fields like:
● Loan/Contract ID
● Borrower/Obligor Name
● Loan Amount
● Interest Rate
● Contract Dates (start, end, payment
dates)
● Equipment Descriptions
● Signatures (presence/absence)
● And many more based on your needs
Each extracted field
shows a confidence score, so you know which values to trust and which to
double-check.
Related Articles
Setting Up Extraction Fields
Define what information to extract from each document type: 1. Go to Queue Settings > Fields 2. For each document type, add the fields you need 3. Specify field names and data types (text, number, date) 4. Configure which loan tape field to validate ...
Exporting Data
Once reviewed, export your processed data: 1. Go to the Exported tab or your reviewed batch 2. Click Export 3. Choose your export format and options 4. Download your files Export Contents: ● Contract-level data sheet ● Asset-level data sheet (if ...
Servicer / Data Provider
Some users primarily provide the ground truth data (loan tape) that ABDP validates against. These users can: ● Upload loan tape files ● View validation results ● Confirm data accuracy
Reviewing Extracted Data
For each document, review the extracted field values: 1. Click on a document 2. See the original document image alongside extracted fields 3. Each field shows: ● The extracted value ● The confidence level ● The source location in the document ...
Batch Statuses
As you work, batches move through these statuses: Status What It Means Pending Documents uploaded, waiting to be processed Processing AI is currently classifying and extracting data Extracted Data extraction complete, ready for review Reviewed You've ...