SAP SuccessFactors introduced AI-assisted proficiency ratings to bridge the gap between inferred skills and formal competency tracking. This feature utilizes Generative AI to recommend specific proficiency levels for skills identified within the Learning module, ensuring that automated skill discovery translates into actionable talent data. For solution architects, the implementation requires a clear understanding of how these recommendations interact with the Talent Intelligence Hub (TIH) and the associated consumption model for AI units.
Alignment with Talent Intelligence Hub
The primary function of AI-assisted ratings is to provide a recommended proficiency level for skills that the system has inferred. While the recommendation is generated by AI, it is not arbitrary; the Gen AI recommendations are strictly aligned to the TIH rating scale configured in your environment. This ensures that a rating suggested in the Learning context remains consistent with the proficiency definitions used across the broader Talent Intelligence Hub.
If the AI identifies a skill that does not currently exist in the TIH attribute library, the system maintains a governance layer. A TIH administrator must verify the new skill before it can be formally tagged for use within SAP SuccessFactors Learning. Once the skill is verified and the AI-assisted proficiency feature is active, the Gen AI will then suggest the appropriate rating based on the established scale.
Managing AI Unit Consumption
Architects must account for the cost model associated with these intelligent features. AI-assisted ratings are classified as Premium AI features, which means they consume AI units as defined for that specific feature. Unlike standard transactional features, the consumption trigger for inferred skills is proactive.
A critical technical detail for budget planning is that AI units are consumed once skills are inferred, regardless of whether an administrator or user subsequently reviews or accepts those ratings. This means that the act of the AI performing the inference and recommendation generates the unit cost, making the initial configuration of inference sources a key lever for managing consumption.
Administrator Implications and Validation
To successfully deploy AI-assisted ratings, administrators should validate the following prerequisites and behaviors:
- Scale Synchronization: Ensure the TIH rating scale is finalized before enabling AI-assisted ratings, as the Gen AI uses this specific scale for its recommendations.
- Skill Governance: Establish a workflow for TIH administrators to review and approve new skills identified by the AI that are not yet in the attribute library.
- Consumption Monitoring: Because units are consumed at the point of inference, monitor the volume of inferred skills in Learning to ensure alignment with available AI unit credits.
- OCN Integration: If using the Open Content Network (OCN), note that AI-assisted skill association only applies to courses sent after the feature is enabled; it does not retroactively tag existing OCN items.
No comments:
Post a Comment