AI-supported field extraction from documents has been available to the market for several years. Early implementations of AI for document processing began to emerge in the mid-2010s. Notable advancements and broader market availability started around 2017-2018, with the development and release of various AI-powered OCR (Optical Character Recognition) and document understanding services by major tech companies like Google, Microsoft, and Amazon.

Today, fully automated extraction of document-specific contents and feeding them into your application or database is a common use case in many industries.

While implementing your own logic to handle one or two simple, consistent, and invariable forms might not be a challenge, adapting to a multitude of different and ever-changing templates and sources quickly poses a complex and costly problem.

What if

• you could simply send your documents to an API from within Ommis and get the requested fields in return?
• adding a new type of document to your recognition stack would only take you minutes?

This topic provides insights on the features and possibilities of integrating Azure Document intelligence into Omnis so that your application can extract the data content out of anything and place it into data fields.

Session 1 – Plattform overview and setup

• Service Blueprint
• Scope and limitations
• Introducing the DocumentIntelligence Plattform
• Setting up the required resources
• Custom classification model
• Custom extraction model
• Training your models

Session 2 – Demo of Omnis integration with Document intelligence

• Providing tokens for accessing the API
• API Tests with Postman
• Creating your own NODE module integrating the Azure SDK with the Omnis Javascript client
• Deploying the module
• Document Classification and Value extraction from Omnis

Appropriate materials and code will be provided as part of the sessions.