Note: A new AI Assisted Reviews experience is available in beta. If your organization has opted in, see What is AI Assisted Reviews? (Beta). To request beta access, see the AI Assisted Reviews: Beta Program page.
15Five uses AI across multiple product areas — Kona, Focus Briefs, AI Assisted Reviews, and Amaya. These are not separate tools. They are connected capabilities that share a common data foundation called the Context Layer.
This article explains how the Context Layer works, what each AI capability does, and how data flows between them. For detailed setup and usage, see the linked articles for each capability.
The Context Layer
The Context Layer is the shared data infrastructure that powers all of 15Five's AI capabilities. It continuously aggregates organizational knowledge into three areas:
- People and Performance Context — A living record of each employee's history within 15Five: roles, tenure, performance reviews, engagement scores, competencies, Objectives, 1-on-1 notes, Check-ins, and key milestones such as promotions or manager changes.
- Work Context — Execution data from connected productivity, messaging, and collaboration tools (such as Slack, Google Calendar, Jira, and Asana), alongside strategic company frameworks like values, goals, and role definitions.
- Added Context — Written guidance provided by users at all levels of the organization to shape how the AI reasons. This includes Company Context (set by HR or Company Admins) and My Context (set by individuals).
Every AI capability described below draws from this same foundation. The Context Layer is managed through the Agents Hub. For full details, see Feature Overview: Agents & The Context Layer.
How data flows through the system
Data enters the Context Layer from two directions: 15Five's own platform data (reviews, Check-ins, 1-on-1s, Objectives, Engagement surveys, High Fives) and connected work system integrations. From there, each AI capability draws what it needs:
- Kona Meeting Assistant joins manager–direct report 1-on-1 meetings, captures structured notes and action items, and syncs them to the 15Five 1-on-1 Agenda. This meeting data becomes part of the Context Layer, where it is available to Focus Briefs and AI Assisted Reviews.
- Kona Coach analyzes meeting data alongside engagement scores, feedback trends, and manager effectiveness metrics to deliver personalized coaching tips to managers.
- Focus Briefs use the Context Layer to generate personalized, recurring briefs for employees and managers. The depth of each brief depends on the admin-configured support level and which integrations are connected.
- AI Assisted Reviews draw from the Context Layer — scoped to the performance period set during cycle creation — to surface evidence, draft review answers, and provide writing guidance. This includes Kona meeting data when available.
- Amaya answers natural-language questions from HR Admins using 15Five platform data, including engagement surveys, performance reviews, Check-ins, 1-on-1s, turnover data, and employee attributes.
Files uploaded directly to 1-on-1s or Check-ins from their respective feature settings pages are not read by 15Five AI. Files uploaded through the Context Layer (My Context or Company Context) are, however.
Data isolation and privacy
All AI capabilities follow the same privacy principles:
- Permission-mirroring: AI strictly mirrors existing system permissions. It never surfaces data to a user who does not already have structural access to the underlying source.
- Private calendar exclusion: Calendar events marked as private are filtered out before ingestion. The AI never sees the subject, description, attendees, or location of private events.
- Admin-controlled depth: Administrators control how much work-tool data reaches each capability through support-level settings. At lower support levels, AI draws only from 15Five platform data. At higher levels, connected work-tool data is included.
- De-identification: Data sent to 15Five's AI sub-processor is de-identified — personally identifiable information is removed before submission.
- No model training: 15Five's agreement with its AI sub-processor prohibits use of customer data for model training. The sub-processor does not retain submitted data.
For full privacy and data handling details, see What data does 15Five AI submit to its AI sub-processor?. For Kona-specific privacy details, see Kona privacy and transcript retention.
AI capabilities at a glance
| Capability | What it does | Who uses it |
|---|---|---|
| Kona Meeting Assistant | Joins 1-on-1 meetings, captures structured notes and action items, and creates a continuous performance record across the review cycle. | Employees, Managers |
| Kona Coach | Delivers personalized coaching tips based on meeting data, engagement scores, and manager effectiveness metrics. | Managers |
| Focus Briefs | Generates personalized, recurring briefs that surface priorities, patterns, and coaching opportunities from 15Five and connected work tools. | Employees, Managers |
| AI Assisted Reviews | Surfaces evidence from the Context Layer, drafts review answers, and provides real-time writing guidance including bias detection. | Managers, Employees (self reviews) |
| Amaya | Answers natural-language questions about people data across engagement, performance, turnover, and employee attributes. | HR Admins |
| AI Synthesized Feedback Insights | Analyzes open-text survey responses to generate themes and summaries. | HR Admins, Managers |
| AI-Powered Recommendations | Generates action recommendations in the HR Outcomes Dashboard based on organizational data. | HR Admins |