Contents
Accuracy issues in unstructured document extraction
Performance drop in upload & prediction for documents with size greater than 1 MB
Performance drop when multiple users access the application concurrently
Multi page table extraction times out for documents with size greater than 1 MB
Masking of documents - No preview option available in the Test page to verify masked image
Lambda & Throttler functions in Richly formatted document extraction need pre-requisite knowledge on Fonduer framework and underlying data model
Separation of machine printed and handwritten cheques
Accuracy issues with Cheque classification
Non-cheque documents are not properly identified
Field extraction accuracy issues for license & PAN documents due to insufficient training data
Supports only PAN & Kerala State Driving License as KYC documents
Slight delay in overall status update when reviewer submits changes
Tilt & Skew correction of images are not available in this release
ServiceNow ticket should have "Type" field set as Ticket Category; for Support UI to show recommendations.
Retraining of ML models to improve recommendations is currently not available.
Associated Conversation Designer project should be re-trained to reflect the changes in workflow mapping
Similarity model is trained based on Boots Ticket dump
Similarity model score varies for small sized Ticket dumps
Periodic ticket dump updates are not affecting the resolution recommendations
Source configuration for monitoring is done from backend
Visualization for a new project is enabled from backend
UI interface for correlation is not available in this release
Correlation fails intermittently due to GRAKN connectivity issue
Contract Management Solution PWF
Accuracy issues in MSA/SOW data extraction
Recalculate option for volume rebates is not available
Addendum details are not available
Delay in processing files & extract data
Volume rebate extraction supports tables with the following 4 columns only – Tier, Minimum, Maximum and % of Level
CEO Name accuracy needs to be improved
Fields selected for extraction cannot be changed
Pre-processor configurations are not applicable for Annual reports
Orientation correction will work only with non-searchable PDF
Accuracy issues with confidence score showing at field level – (eg Incorrect value extracted with high confident score.)
Not possible to select the ‘Waiting For Approval’ documents all together by Supervisor user
Few fields are missing in excels downloaded and created via report generation
Line item window is not movable in Preview Page
Confidence Score is displayed for newly added (via ‘Add Invoice’) invoices which is not relevant
Reassign functionality is not working for assigned invoices
All fields selected in project configuration is not getting displaying in Preview Page
Extraction time for a single page invoice with 1 line item is 142 seconds
Fieldwise extraction accuracy is not good for all 14 fields