Contact us

Mosaic Entity Extractor

Enterprises today are on the cusp of next wave of disruption, triggered by Artificial Intelligence & Cognitive technologies. At such a critical juncture, enterprises should leverage Artificial Intelligence technology to deliver business outcomes to redefine business models and customer engagements.

LTI’s Mosaic Entity Extractor is the state-of-the-art new age framework to process unstructured data. Mosaic Entity Extractor leverages deep learning to mimic the human behaviour to process the unstructured data to extract the required data in an automated way. Information is trapped in the physical documents and images in the enterprise either available as the legacy data in warehouses or being used in the business transactions. Mosaic Entity Extractor enables a new way to deal with the complex unstructured documents with varying formats and layouts by enabling different cognitive functions which can be skilled by business users themselves. Mosaic Entity Extractor overcomes the challenges, of template driven approach of data extraction from documents, which is followed by existing OCR based solutions.Healthcare industry deals with numerous forms for processes like requesting a prior authorization, appeals done at various levels and handwritten letters which can be used to submit a grievance. All such forms need intelligence in data extraction and submission to multiple databases. Typically, these documents are physically read, and the details are typed in an internal system. This is not only a tedious and mundane task, but also has a lot of margin for human errors.

Key features of Mosaic Entity Extractor:

  • Ability to extract data from structured and unstructured sources not limited to documents but also tabular data and images
  • Method to classify documents based on contents or visual characteristics
  • Ability to train through learning models when new fields are discovered in the extraction process
  • Capability to integrate with OCR engines and NLP tools providing multi-language support

Mosaic Entity Extractor for Healthcare
LTI’s Mosaic Entity Extractor classifies, reads and analyzes images and documents using its deep learning-based AI models. With availability on-cloud and on-premise, it has integration capabilities with OCR engines such as Google Vision and Abby FineReader SDK. Healthcare processes uses form-based input in multiple processes like prior authorization, claims submission, appeals and grievances.

AI based data extraction from multiple types of paper claims (CMS-1500, CMS-1450), prior authorization forms, appeals & grievance for different levels can be best done using Mosaic Entity Extractor. The data extracted is routed through a workflow-based process leveraging Human-in-the-loop (HITL) feature or verification.


Reach us

I agree to receive communication from LTI.
Refer LTI privacy policy to know more about how we maintain privacy about your data.


Our Success Stories

Employee Benefit platform

LTI developed cloud-based Employee Benefit Platform for a Life & Health Insurance Broker

Event Based Intelligence

LTI developed complex algorithm and Event-based Intelligence for Major US-Based Hospital

Health and Wellness

Health & Wellness Platform Implementation for Singapore-based Life Insurance Company

Intelligent Automation

LTI implemented Intelligent Automation for US-based Healthcare Organization

Workflow Implementation

Workflow Implementation for Life & Health Insurance Company

Application Maintenance

Application & Infrastructure Maintenance for US-based Medicare Health Insurance Organization