AI-powered document classification service

The service achieved a 98% accuracy in classifying both hand-written and typed documents.

Client

Project and document management systems (DMS) provider offering software products for construction and real estate industries.

Objective

Streamline and simplify manual document entry process for the end users of the client’s DMS.

Solution

Our data science experts reviewed samples of the typical documents that get entered into the system and built the Machine Learning model that broke them down into categories based on the OCRed content. Then the software engineering team wrapped the model into a group of APIs deployed to the AWS cloud and integrated the solution with client’s document management system.

The service offers the following features:

Parsing of the typical construction and real estate documents in DOC, XLS, PDF, PNG, JPEG, BMP and AutoCad’s DWG formats
Recognition of both hand-written and typed text
Classification and labeling of the documents based on the predefined rules
Integration with DMS platforms
GDPR compliance

Result

Our cloud-based solution achieved 98% accuracy in the classification of both hand-written and typed documents. It significantly reduced the manual burden on the end users and brought the client’s software to the next technological level powered by AI.

Accuracy enhancement: the service achieved a 98% accuracy in classifying both hand-written and typed documents.
Operational efficiency: significantly reduced the manual burden on end-users, enhancing overall operational efficiency.
Technological advancement: elevated the client’s software to the next technological level, empowered by AI.