The employeeInsite platform incorporates artificial intelligence technologies to enhance user experience and provide intelligent assistance. AI is used in two primary ways:
- AI-Powered Chat Assistant - An intelligent conversational assistant that helps users navigate the platform and answer questions about their organization's data
- Report Generation - AI-powered generation of personalized reports after assessment completion
All AI features are designed with privacy and security as top priorities, ensuring that sensitive employee data is handled appropriately, and user privacy is protected.
AI Features
Chat Assistant
The AI Chat Assistant is an intelligent conversational interface that helps users address leadership challenges and learn leadership techniques utilizing organizational context along with individual motivational data (non-confidential data only). No engagement data is used to assist in answering responses except that which has been anonymized and aggregated (similar to what can be seen in data reporting dashboards).
- Answers questions related to leadership and employee motivation utilizing the employeeInsite framework to offer solutions grounded in leadership and employment best practices.
- Offers contextual assistance based on user's role and access level.
- Utilizes state of the art defensive prompting techniques to reject any queries outside of its scope of expertise.
Report Generation
After users complete assessments, AI is used to generate personalized reports that provide insights and analysis based on their responses. Reports include personalized insights, actionable recommendations, visualizations, and contextual analysis tailored to each user's assessment results.
AI Providers
The platform works with leading AI providers to deliver high-quality AI features:
Provider | Description |
OpenAI | Enterprise-grade AI services with strong performance and reliability |
Google | Advanced AI models with high-quality outputs and scalable infrastructure |
Anthropic | AI safety and reliability focus with advanced reasoning capabilities |
OpenRouter | AI model routing and management for access to multiple models |
Data Usage
The platform utilizes and stores data from various sources to ensure AI outputs are contextually relevant.
Data Source | AI Usage |
Employee HR Data | Some HR data is used to provide context to AI models. This may include an employee’s job title, pay type (hourly/salary), employment status (full-time/part-time), gender, generation (Millennial, Gen X, etc.), department, and approximate organizational tenure.
Other information is optionally collected to power aggregated and
anonymized reporting (i.e. exact age/birthdate, compensation, exact
tenure, etc.) but this information is not provided to AI providers. |
Organization Structure | Information related to reporting and team structures is used to provide context to AI models. |
Employee Preferences Data | Motivational and preference data provided by employees via the employeeInsite assessment is used to provide context to AI models |
Employee Job Satisfaction Data | Job satisfaction data is only provided to AI models in an aggregated and anonymized way. |
Data Privacy & Security
Data Handling Principles
- Minimal Data Exposure: Only necessary data is sent to AI providers (see above table).
- Access Controls: AI features respect user access levels. Organization-level data isolation is maintained and role-based restrictions are enforced.
- Data Isolation: Each organization's data is isolated. Cross-organization data leakage is prevented and multi-tenant security is maintained.
What Data is NOT sent to AI Providers
- Passwords or authentication credentials
- Financial information
- Complete employee records (see Data Usage table for a breakdown of what is provided)
- Unauthorized organizational data
- Data that would be unacceptable if it were directly accessed by the user
User Controls & Transparency
- Users can start new conversations and delete conversations at any time
- Users control what questions they ask and can stop conversations
- Clear indication when AI is being used with real-time streaming of responses
We understand that AI is not perfect and with enough persistence a user can theoretically hack the AI to gain access to any resources that the AI has access to. While we take preventative measures to prevent this from occurring, all tools that the AI has access to are scoped to apply the same role and resource-based access controls that the user making the request has (making it impossible for a user to “hack” the AI to get access to data that they wouldn’t otherwise already have access to).
Monitoring & Compliance
The platform includes comprehensive monitoring including usage tracking (number of AI requests, response times, error rates), security monitoring for unusual usage patterns and potential abuse detection, and logging for auditing purposes. All AI interactions are logged with timestamps, user identification, request types, and success/failure status.
For questions about AI usage, privacy concerns, or feature requests, please contact our support team (support@employeeinsite.work).
Powered by