Document Indexing, Data Extraction and Data Validation

Document Management Solution

VisiCapture is an industry-leading auto-indexing and data extraction solution. It offers a high level of accuracy and quicker processing time using best-of-the-breed data capture techniques. VisiCapture is hosted on Amazon cloud and is available to you as a software-as-a-service model on a transactional-based pricing model.

Visionet currently supports more than a thousand cloud-based users who are processing millions of documents every month. Visionet has also perfected the free form text based indexing and data extraction for various industries like mortgage, retail, insurance, logistics and healthcare. No new hardware or software is needed to benefit from our years of specialized experience with the following simple steps:

  1. Upload the scanned documents though a secure connection to Amazon Cloud
  2. Index the file into as many buckets as desired
  3. Capture relevant data in desired format
  4. Download indexed documents and data or securely share it with partners

At the backend, VisiCapture engine takes a raw imaged document file, converts it into a proper format, indexes the pages into the documents and desired sub-packages and uploads the correctly categorized documents into client’s repository. VisiCapture can also convert the document images into data, and categorize that into searchable data fields. This information can also be validated as per the pre-defined process and the business needs. The entire process takes couple of minutes saving precious time and money while improving accuracy.

Key Features:

  • Automates document indexing in a pre-defined order
  • Auto recognition of documents based on multiple strategies
  • Auto extraction various types of data from recognized documents
  • Auto document validation based on the pre-defined rules and processes
  • Batch processing allows to schedule large volume of files from folders
  • Indexed documents, data and reports are available through a secure portal
  • Includes API based integration with either XML based or table based output
  • Auto template coordinate alignment for full page analysis and zonal recognition
  • Auto corrects scanning errors like skew or speckle by de-skewing automatically
  • Easy-to-use interface for creating pre-defined templates, and improve recognition
  • Manages exceptions based on user’s inputs or on pre-defined corrective actions
  • Integrates with existing document imaging systems and scanning infrastructure
  • Supports a broad range of image file types, image resolutions and output formats
  • Delivers high accuracy for document conversion and archiving
  • Is scalable and can easily cater to large data volumes


VisiCapture offers configurable document indexing schemes that automatically recognizes different types of documents in a file.

Let’s take a mortgage file as an example – VisiCapture automates the indexing as per the pre-defined mortgage processes, and all the mortgage documents in a loan file are indexed including the multiple copies of the note, which are identified and tagged by date. If you have paper files, we can scan those and make these available through a secure portal for you to review and download.


VisiCapture recognizes scanned document and converts these into searchable data. It intelligently stores the desired data fields in a manner that it becomes information.

For example, all mortgage related documents are duly recognized, relevant data fields, the mortgage clauses and other textual information is also extracted during this process, in such a way that all standard data fields are available for underwriting, loan reviews and any other pre-defined business processes.


VisiCapture verifies completeness and integrity of documents. It works intelligently to recognize the document from the uploaded image files, and automatically checks whether the information is consistent across all the documents.

For example, it automatically recognizes the file type say a loan file and validates whether the appended documents have the same borrower name and the property address. It also validates whether the documents are signed and are signed

For more information about how Visionet’s data capture solution and services can help your business save time, reduce operational costs and improve data accuracy, contact us today.

Use Cases:

HUD-1 Preparation

VisiCapture is being used to auto generate HUD-1 from the scanned data based on the Closing Instructions. The scope of this assignment involves 250 data fields, comprising of 4-pages, and requires field and even character level confidence on data collection – all this with a large volume of thousands of forms to be processed every day, and provide dashboard reporting to the executive management for monitoring the progress.

VisiCapture offers more than 99% accuracy rate on average, whereas at least 90% of these documents are captured with 100% accuracy rate. This entire process takes less than a minute. VisiCapture is capable of processing multiple documents simultaneously through multithreading.

Invoice Processing

VisiCapture supports invoice data extraction for millions of pages every month, and this usually comprise of data volumes in terabytes from the scanned images. With it ability to process thousands of documents a day with say 1.65 million pages with over 50 million rows – it provides a very high accuracy and completes this kind of volume in a relatively short time by using multi-threading approach for processing such batch jobs on highly reliable servers.

Post Closing Review

VisiCapture is being used by hundreds of lenders and servicers and for indexing and validating over a set of over 50 different documents in a loan file for post closing review process. With over 200 pages, it utilizes state-of-the art capture techniques to identify documents and extract data from the documents – all within few minutes.

The data fields are then used for validation, and hundreds of documents are processed a day. VisiCapture offers more than 96% accuracy rate in indexing, and above 90% in the document validation. The average time for processing a loan file on its post closing review is less than 5 minutes.


Additional Resources



Challenges in meeting the GSE mandate for uniform closing dataset (UCD) delivery
Date: September 20th, 2016 02:00 PM EST

Learn More & Register