Solutions

A Modifiable Intelligent Forms Processing System

The Problem

Manually reviewing a high volume of paper and digital forms

High rate of errors associated with the manual review of forms

Manually classifying forms for specific business needs

Entering data from forms into an enterprise system

Dedicating time, resources, and money to manually process forms

Intelligent Forms Processing System

Specto augments the processing of paper and digital based documents, integrates seamlessly into any workflow, and can be modified based on client needs. Specto extracts and classifies data from incoming documents (digital or paper based) and applies the output to downstream systems and workflows. 

Specto uses Artificial Intelligence technologies that include supervised machine learning (ML) techniques to continually improve document classification and data extraction. Character Recognition (CR) allows for data extraction (unstructured and structured) from typed text. Additionally, all the data from the system can be exported to different file formats, and then to databases, or systems. More importantly, all the AI technologies NeuroSoph uses are secure.


Use Cases

Augment Document Processing Workflows

In 2017 just as in 1917, government employees spend huge a amounts of time on paperwork. A recent survey of state and local officials found that 53 percent had trouble getting their work done in a 35-to-45 hour week due to excessive paper work burdens. A government agency could use Specto to augment the processing and classification of forms, reduce backlogs, and reduce frustrating constituents and employees.

Eliminate Historical Backlogs

Digitized forms and documents contain valuable data for business and organizations however they may not have the time or resources to have staff manual go review the documents and enter the information into a system or database. For example an entertainment company may have thousands of historical digitized record contracts the entertainment company could use Specto to read the thousands of contracts, extract information with Optical Character Recognition, create a searchable index or store the information in a system or database.

Prioritization of Forms 
An insurance firm requires a large number of incoming invoices and claims to be processed promptly and efficiently in order to speed up the payment process. Decisions need to be made on which invoice streams to prioritize for processing, and how this should be done. Invoices are categorized using Specto. Specto determined the form/invoice type, the category to which the document belongs and to which employee the document should be transferred.
 

How it Works

Benefits of Specto

Increase speed of processing document due to automation

Improves the accuracy of document classification and increases operational efficiency

Reduces the error rate associated with the manual review of documents

Frees up internal resources to work on more creative tasks and focus on the mission

Enables internal resources to improve the customer experience

Improves the quality and security of data


Form Design

NeuroSoph can assist with designing forms so that Specto can process forms with higher efficiency and accuracy. NeuroSoph offers deliverables across:

  • Designing forms from “scratch” based on the information that needs to be collected
  • Advising how to display form elements such as check boxes, radio buttons, text areas, and more
  • Recommending how to improve existing form designs

Features

Allows the users to monitor system progress and performance, verify information, and tag forms that need verification.

Scanned documents from different sources (printers, scanners, fax machines, etc.) often exhibit distortions – including scaling, rotation, and blurring. Specto uses image processing techniques to clean up these images, so they can be processed and analyzed accurately.

Using supervised machine learning Specto can automatically classify documents into specific categories based on client requirements.

Specto can recognize a full form/document by detecting all the page numbers and associated objects such as form fields (e.g. Address, Name, Date), text boxes, and handwriting.

Data is automatically extracted from paper or digital documents structured and unstructured using Optical Character Recognition (OCR).

The user can view all the forms that Specto processed and any of the extracted data. Specto automatically assigns a quality score to the form processed and information extracted. If a document has a low-quality score it is sent to a user for review.

The user can tag documents to give feedback to the machine learning model. When enough feedback is collected NeuroSoph will retrain the model. Over time, the user’s verification of the data makes Specto smarter and allows for rectification of edge cases, error files, unmatched data, and duplicates.

Specto gives the user the ability to export to CSV, JSON, and XML this allows the processed forms to be loaded into a client database.

Each Specto instance creates a storage database, enabling the client to have a history of all the data Specto has processed and the ability to run analytics and reports.

Specto can identify if the form is in the right order and if not, it will intelligently reorder the form if the pages are misaligned or pages are missing.

Automatic engine recovery feature that senses when an engine fails and automatically re-initializes it for the next form.

Future Features

Extracting handwritten text using Intelligent Character Recognition (ICR).

Document summary techniques, condensing forms that are long into a single page (or paragraph) allowing long legal documents, government forms, and proposals to be quickly paraphrased.

Flexible Pricing Models

Project based, licensing, subscription or transaction based.