Our Approach to AI
At Sylla, we believe in the responsible and transparent use of artificial intelligence to enhance educational experiences. Our platform leverages AI technology to support educators while maintaining the highest standards of privacy, security, and ethical considerations.
AI Implementation
Our platform utilizes two primary types of AI processing:
User-Generated Content Processing
- We employ Anthropic's Claude models for expanding course topics and generating relevant search queries
- These models do not train on user input/output data, ensuring your content remains private
- All processing is done with explicit context boundaries to prevent bias and maintain relevance
Content Analysis
- We use VoyageAI's voyage-large-2 embedding model for processing educational resources
- This enables semantic understanding of content while maintaining data privacy
- No user or institutional data is used for model training or enhancement
Data Protection and Privacy
We maintain strict data protection measures:
- No institutional data is integrated into external AI models
- User interactions are not used to enhance or retrain AI models
- Module information is processed solely for operational purposes
- All data processing complies with relevant data protection regulations
Transparency and Safety
Our commitment to transparency includes:
- Clear visual indicators when AI features are in use
- Distinctive interface elements for AI-generated content
- Explicit marking of AI-powered features and suggestions
- Built-in safety features and content filtering through Anthropic's Claude
- System-level guardrails to ensure educational appropriateness
Educational Integrity
We prioritize academic integrity by:
- Maintaining clear connections between source content and AI-generated suggestions
- Implementing careful constraints on AI responses for educational contexts
- Ensuring all AI-generated content aligns with academic standards
- Supporting proper educational practices and pedagogical goals
Accessibility
Our AI features are designed with accessibility in mind:
- Screen reader compatibility
- Configurable interface options
- WCAG guideline compliance
- Clear, readable presentation of AI-generated content
Institutional Rights and Protection
We respect and protect institutional interests by:
- Not restricting the use of our solution within AI tools and services
- Ensuring full accessibility to all authorized users
- Not introducing new institutional liabilities
- Protecting against intellectual property infringement
- Maintaining strict access controls for third parties
Continuous Improvement
We are committed to:
- Regular monitoring and evaluation of AI performance
- Updating our safety measures as technology evolves
- Incorporating user feedback to enhance our AI features
- Maintaining transparency in our AI implementation
Contact
For questions about our AI policy or to report concerns, please contact our team at info@sylla.io
This policy is regularly reviewed and updated to reflect the latest developments in AI technology and educational needs.
.svg){kind=logo}
