Problem Summary
Upgrade, Inc. completes client questionnaires year over year that contain largely the same questions. In many cases, answers remain unchanged from previous years. To improve efficiency and answer reuse, they want more precise control over which past project AI Assistant and Answer Library Search reference.
Specifically, they want the ability to filter by a specific past project name at the Project Settings level, rather than relying on broad or indirect filters.
Current Functionality
1. Recommendation Engine – Past Project Search (Question-Level)
- Users can filter by specific past projects by name at the question level.
- Filters can optionally be frozen for a section.
- This search looks across all past projects in the client’s instance, regardless of whether content was saved to the Answer Library.
- This is the behavior the client finds most useful and wants replicated at a higher level.
2. Project Settings – Answer Library Search
- Users can filter by Project Name.
- This filter only searches projects with content saved to the Content/Answer Library.
- Projects without saved library content are excluded, even if they are relevant.
3. Project Settings – Past Project Search
- Users can filter past projects by:
- Project status
- Project created date
- Users cannot:
- View a list of past projects by name
- Select a specific project to scope results
Gaps & Limitations
- There is no way to scope AI Assistant or Answer Library Search to a single, known past project by name at the project level.
- Existing project-level filters are either:
- Too broad (status/date), or
- Too restrictive (Answer Library–only projects).
- This forces users to rely on question-level filters or manual review, reducing efficiency for repeat questionnaires.
Requested Functionality
Past Project Name Filter in Project Settings
Add a Past Project Name filter in Project Settings that:
- Mirrors the behavior of Recommendation Engine Past Project Search (question-level).
- Searches across all past projects in the client’s instance, not just those with Answer Library content.
- Allows users to:
- View and select one or more past projects by name
- Scope results holistically at the project level
Scope & Usage
- The selected past project(s) should be used consistently across:
- AI Assistant
- Answer Library Search
- The filter should apply globally to the project, not just individual questions.
Value
- Improves accuracy and relevance of AI-suggested answers
- Reduces manual searching for repeat questionnaires
- Aligns project-level behavior with an already successful question-level pattern
- Supports customers with annual or recurring security/compliance workflows