We’re excited to announce that the latest update is now available for the KnowledgeLake Cloud.
Feature Update Overview
(some features listed below may be add-ons)
|
 |
KnowledgeLake Intuitive AI
KnowledgeLake's Intuitive AI is the next generation of Intelligent Document Processing (IDP). This new technology transcends automatic classification and extraction and now enables true document understanding. This state-of-the-art capability enables the platform to classify and extract critical information from documents effectively, without the need for preliminary training. This approach, known as 'Zero-Shot' IDP, not only delivers highly accurate results but also offers the flexibility to refine these outcomes over time. Users can effortlessly tune and improve the training set through a user-friendly no-code/low-code interface, ensuring continual enhancement of the system's performance and accuracy. With these updates to IDP, you can expect marked improvements to our OCR accuracy. Not only that, but we are also excited to announce handwriting (ICR) recognition and extraction. There is no additional configuration or training needed for ICR, it is now accounted for when OCR is performed.
|
 |
Enhanced Upload Experience
We are excited to announce the introduction of Upload Profiles, a significant new feature that allows users to create and modify shared settings through a streamlined interface, enhancing both the Upload App and Office Add-Ins experience. Key updates include a new Upload Settings page with categories for Office Add-Ins, Upload Restrictions, and the ability to assign specific workflows, icons, and access permissions to each profile. These improvements are designed to optimize document uploading processes, making it more intuitive and efficient for all users.
|
Improvements to Workflow
-
-
Capture Directly to a Workflow – The Scan app, Upload app and Import jobs now support releasing directly to a Workflow. This removes the need to go through Process Designer before sending captured documents to your desired workflow.
-
Custom Events – You can now write the exact events needed, including metadata values at any point in a workflow. This minimizes the number of sources, such as SQL, needed when compiling a report.
-
Events Filter – In order to reduce the volume of events being stored in a Secondary Event Provider, administrators may now filter out events from specific sources throughout the platform such as when a job runs, a user makes a change in an app and many more event types.
-
New Functions in Custom Step – The Custom Step in workflow has been updated to include all new functions for editing the structure of Documents and pages in a work item. You can now create documents, delete documents, shift or copy pages from one document to another and much more.
-
Make Web Service Calls In Workflow – We have added even more flexibility to workflow, allowing you to easily work with the APIs of external systems. This is ideal for getting or setting data in those systems right from within the KnowledgeLake workflow.
|
Other Improvements
-
-
New Form Options - Two new form options have been added: matrix and panels. Matrix controls enable you to gather survey-style data on the form. Panels enable you to do repeating sections such as gathering a list of personal information, adding entries to a timesheet and other cases where the number of expected entries is open-ended.
-
Rearrange Gizmos in a Task Layout - Gizmos can now be rearranged when designing a Layout, reducing rework when a change is needed.
-
Download PDF in Scan - Users can now download a copy of a document directly in the Scan app as a PDF file.
-
Local Tables - With Local Tables, users can effortlessly create and manage two-dimensional tables, thanks to a new source type and an intuitive table editor, complete with useful hotkeys for streamlined operations.
-
Fuzzy Index Matching - Fuzzy Matching introduces an innovative approach to auto-indexing, allowing for flexible string comparisons that correct minor discrepancies, ensuring more accurate and efficient data matching
|
Comments
0 comments
Article is closed for comments.