by Sujith Parakkunnath, on July 13, 2022 10:00:00 AM PDT
We entered 2022 with focused goals, an industry-driven product roadmap and a futuristic perspective. As we kick off this third quarter, we remain focused on providing you features to help our users reap the best results from our Intelligent Document Processing (IDP) platform.
Here are some of the major features and enhancements from our latest release.
Business Rules for Document Processing
Infrrd's IDP solution now provides options in the user interface to perform derivations or calculations on extracted data to transform them based on your individual requirements. Users can achieve this by configuring business rules. Previously, contacting Infrrd’s customization team was required and now, some of these capabilities are available from the IDP user interface. This will result in significant time savings as available business rules can simply be configured and managed within the Infrrd IDP platform rather than routing it through the Infrrd customization team.
Infrrd Knowledge Graph
Additionally, with this new release, new document models can be created by re-using our knowledge graph, a repository of all the data types that our AI already understands. This collection of more than 500+ fields can be used to create a new model for a document that our system has never seen before — without any training.
Table Extraction Enhanced
Table extraction, including the user experience, also has been enhanced. Now, Infrrd IDP users have options to extract the Prominent Table. The Prominent Table is identified by the IDP system as the most important table in the document. Users also now have the ability to extract all the tables in a document. Additionally, to ensure an enhanced user experience, tables are now listed in a separate tab after extraction or during a correction. A new deep learning architecture is used for training table structure, which is expected to improve the accuracy of row and column identification.
Google Drive Integration Launched
In Infrrd’s IDP, users can now upload documents to a model from a Google drive by configuring the data source. This allows documents to be uploaded to a model easily and securely for the extraction process. Documents in the configured Google drive will automatically be uploaded to the model, based on a configured polling interval. Instead of uploading documents to Infrrd’s system, now users can just place them in their Google drives. Infrrd will pick up the documents from users’ Google drive in regular polling intervals and place the results back in there.
