-
AI Draft – Enforce character limits by rejecting or flagging responses before acceptance
Currently, when character limits are configured in AI Draft (at the project AI instruction level, section level, or via AI Draft instructions), the AI does not reliably honor those limits. Responses that exceed the configured limit are still surfaced to the user and can be accepted. The system only flags the violation with…
-
AI Draft Should Adhere to Character Type and Character Limits
I’m submitting this enhancement request on behalf of our strategic client Qualatrics. Description of Issue: Qualtrics discovered that when importing a project and setting limits on character type or character/word limits, AI Draft does not recognize or comply with these configured restrictions. As a result, Proposal…
-
Improve AI Draft Response Accuracy for Parent-Child Questions
Enhancement Request: Improve AI Draft Response Accuracy for Parent-Child Questions Problem Statement Customers have reported scenarios where AI Draft generates responses for a broader parent question instead of the specific child question. This behavior occurs when questions contain special formatting or characters,…
-
Bulk Revert Functionality in AI Draft Review Answer Workflow
Ticket Reference: TKT-217460 Customer: Nasdaq Requester: Kritika Gour Enhancement Summary: The customer has requested an enhancement to improve the usability and efficiency of the AI Draft Answering Agent’s “Review Answer” workflow. Current Behavior: At present, the “Revert” option is available only at the individual…
-
Direct Navigation and Inline Editing for “Answer Not Found” Questions in AI Draft
Customer: Nasdaq Ticket Reference: TKT-217454 Requirement Summary: The customer has requested an enhancement to improve the usability and navigation workflow for questions marked as “Answer Not Found” during the AI Draft process. Current Limitation: When AI Draft is unable to generate responses for certain questions, those…
-
Advanced Search/AI Filters - Display only tags with content under them
Current Issue: It was confirmed that it is expected behavior for tags that do not have content under them (i.e. zero counts) to display as selectable options within advanced search (content library & project) + AI Assistant Data Filtering. Only ASK and Quick Filters/Facets limit display to tags with content under them.…
-
Section-Level Filter Support in AI Draft – Review Answer Page
Summary: Users have requested enhanced filter granularity within the AI Draft – Review Answer page. Currently, filters applied in the Review Answer interface are applied globally across all sections in the project, which limits the ability to perform targeted review and analysis at a section level. Current Behavior:…
-
Add a prompt editor within the AI Draft workflow
Enhancement Request Summary: Add a prompt editor within the AI Draft workflow to support iterative prompt refinement and response regeneration without requiring users to discard and restart the draft generation process. Business Requirement / Use Case: Currently, users are unable to edit or refine prompts once AI Draft…
-
AI Draft “AI Response Generation Stalled” error
Daria (daria.lysohub@securiguard.com) from Securiguard raised a concern that AI Draft throws the error “AI Response Generation Stalled” even though the response contains placeholder/help text, which seems like odd behavior. The customer mentioned that they have placeholder/help text in their source file within the project,…
-
Answers completed by AI draft not appearing as completed answers but rather completed by AI draft
Amadeus Hospitality have requested: Could we review updating answers completed by AI draft as not completed but perhaps to a different colour or status so that they weren't completed answers, rather they were completed by AI draft. Teams will still need to review those answers and they wanted a way to differentiate between…
-
Question level filtering for AI draft - Cognite
The Problem Currently, AI Draft apply a library filters to an entire project/sectios. Large clients like Cognite often require segmented sources for a single RFP (e.g., 10 questions from the Technical Collection and 10 from the Legal Collection). Current Workaround: Manual, repetitive editing or running multiple drafts,…
-
Count AI Draft as AI Usage
Currently if you run AI Draft in a project and the results are pulled directly from the content library (no answer generation) the reporting considers this as "Answer Library Used" without mention of AI. Voya believes this skews the AI usage reports and would like to see a designation like AI Draft - Answer Library Used so…