The sub-processor does not use 15Five customer data to train its model.
Key Rules
- Engage feedback insights: Open-text survey responses are submitted along with 15Five's driver definitions and survey statements. The LLM uses this to associate feedback with a specific engagement driver.
- AI-assisted reviews: Employee review resources (High Fives, wins and challenges, feedback), self-reviews, and peer reviews are submitted and summarized as Strengths and Opportunities. When a manager prompts the LLM to draft or improve an open-text response, objectives and 1-on-1 data are also included.
- Kona: Meeting notes are encrypted at rest and in transit. Notes are accessible only to the manager and their direct report.
- Data is sent from the 15Five web application server to the LLM sub-processor entirely within the United States.
- The sub-processor does not retain data after responding to a query.