TPRM issue recommendation skill
The issue recommendation skill in Now Assist for Third-party Risk Management (TPRM) uses generative AI to suggest potential issues based on historical assessment data, helping assessors and reviewers identify relevant risks more efficiently while retaining control over which issues are created.
Overview of issue recommendations
Third-party assessment reviewers [sn_vdr_risk_asmt.vendor_assessment_reviewer] can use TPRM issue recommendations review completed third-party assessments after questionnaire responses are submitted. The TPRM issue recommendation skill analyzes historical issues created from all assessments, along with historical Smart Assessment and Classic assessment questions and responses, to generate suggested issues when semantically similar questions exist and the intended meaning of the answers is consistent.
For more information, see Generate issue recommendations for TPRM and Create or dismiss issues using recommendations.
Benefits of issue recommendations
- Generates potential issues based on intention from historic questions and answers.
- Helps reviewers focus on relevant risks without reanalyzing every assessment response from scratch.
- Supports consistent issue identification across all assessments while preserving human judgment.
Components of an issue recommendation
When issue recommendations are generated, each recommendation includes contextual information to help reviewers evaluate its relevance:
- Issue name and description
- Related smart assessment question
- Assessment type used as reference data
- Priority derived from historical issue data
- Historical issues used to generate the recommendation
| Element | Description |
|---|---|
| Issue description | Summary of the recommended issue generated from historical assessment findings. |
| Smart assessment question instance | Assessment question associated with the recommended issue. |
| Types | Source of the historic data. By default, it is from the previous assessments. |
| Historical issues | Previously created issues that were used as reference data to generate the recommendation. |
| Priority | Priority of the historical issue. |
The accuracy of issue recommendations depends on the quality and configuration of Smart Assessment Engine templates, including supported question types and historical issue data.