1. Introduction
This Privacy Policy outlines how Kriyo ("we," "our," or "us") collects, processes, stores, and safeguards information when users access and utilize Kriyo AI Services. As artificial intelligence technologies require data inputs to generate meaningful results, we are committed to maintaining the highest standards of data protection and transparency. By using Kriyo AI Services, you understand and acknowledge that your data will be handled in accordance with the practices described in this policy.
2. Information We Collect
To provide and maintain high-quality AI functionalities, Kriyo collects various types of data resulting from your platform interactions. This collected information includes:
- User Prompts: The text, queries, data inputs, and instructions you manually enter or submit into the AI interface.
- AI-Generated Outputs: The text, responses, summaries, and content generated by the AI models in response to your prompts.
- User Feedback: Direct assessments, ratings, or commentary you explicitly provide regarding the quality of the AI outputs.
- AI Interaction History: Logs of your conversations, time stamps, and historical usage patterns within the AI features.
- Usage Analytics: Technical data regarding how frequently and effectively different AI tools are utilized across the platform.
- Device Information: Metadata concerning the hardware, operating system, web browser, and network settings used to access the services.
- Feature Usage Information: Metrics capturing which specific AI capabilities, models, and tools are accessed during a given session.
3. Purpose of Processing & AI Model Improvement
Kriyo processes the collected information for specific, legitimate business and operational purposes. Specifically, this data is utilized to:
- Deliver AI Services: Execute your requests and generate the precise content or insights you ask the system to perform.
- Improve AI Quality: Refine underlying algorithms to ensure more contextually accurate and relevant outputs for educational environments.
- Optimize Prompts: Enhance system-level engineering to help the AI better interpret user intent.
- Train Future AI Models: Teach subsequent generations of machine learning models how to respond more effectively to industry-specific workflows.
- Evaluate AI Performance: Audit system accuracy, identify systematic errors, and benchmark platform performance.
- Develop New AI Capabilities: Identify operational gaps and build entirely new AI-driven tools, features, and automations.
- Improve Customer Experience: Customize the user interface, streamline user workflows, and provide tailored technical support.
- Maintain Platform Security: Monitor for abusive behavior, prevent fraudulent activity, and secure the overall infrastructure against cyber threats.
- Privacy Safeguard for AI Training: Information is retained long-term to train future machine learning models or evaluate AI quality, Kriyo applies strict pseudonymisation, de-identification, or data masking protocols to ensure your data is scrubbed of identifying details before processing.
4. AI Processing
To deliver advanced, conversational, and analytical intelligence, your AI requests and prompts may be routed through and processed by industry-leading external AI infrastructure providers. These third-party vendors, which include OpenAI, Google Gemini, Anthropic Claude, Mistral and other specialized AI infrastructure vendors selected by Kriyo, act as data processors. They are contractually obligated to process the submitted information solely for the purpose of executing the requested AI functionality and are strictly prohibited from utilizing this data for external or independent purposes.
5. PII Protection
Kriyo prioritizes user privacy, especially within educational contexts. Before any prompt or text input is transferred to external third-party AI providers, Kriyo deploys advanced, automated Personally Identifiable Information (PII) masking technologies. This system scans your input in real time to automatically identify and mask selected sensitive data points.
Examples of automatically masked information include:
- Full or partial personal names.
- Telephone and mobile numbers.
- Personal identification numbers.
- Student IDs.
- Financial data, including bank accounts.
Furthermore, Kriyo enforces a strict data policy ensuring that we do not intentionally use unmasked personally identifiable information for the training or refinement of any AI machine learning models.
6. Human Review
Authorized Kriyo personnel may perform manual human reviews of specific AI interactions. This human intervention is conducted only when strictly necessary, such as fulfilling explicit user support requests, conducting security and abuse investigations, verifying quality assurance standards, optimizing product functionality, or auditing AI compliance. All human review sessions are subject to strict internal logging, data minimization principles, and are accessible only by authorized personnel bound by rigorous confidentiality agreements.
7. Data Retention
Kriyo retains AI-related information for the duration necessary to fulfill the operational goals outlined in this policy. This includes maintaining records for core service delivery, defending platform security, satisfying statutory compliance, running aggregate analytics, refining product workflows, and facilitating machine learning research and development. Because different types of data serve different operational functions, precise retention periods vary based on specific legal obligations and operational requirements. Once data is no longer required, it is securely deleted or permanently anonymized.
8. Security
Kriyo implements a robust framework of technical, administrative, and organizational security controls designed to protect AI-related data from unauthorized access, loss, alteration, or disclosure. Our multi-layered security stack features industry-standard data encryption, strict identity and access management controls, automated PII masking filters, continuous real-time security monitoring, and highly restricted employee data access permissions.
9. International Transfers
Kriyo AI Services leverage a globally distributed cloud infrastructure. Consequently, the information you submit may be transferred to, stored in, and processed across multiple geographic jurisdictions where Kriyo or its trusted third-party AI infrastructure providers maintain operations. When transferring information across international borders, Kriyo implements legally approved and appropriate data transfer safeguards to ensure your personal data remains fully protected under equivalent data protection standards.
10. User Consent
By accessing, enabling, and actively using Kriyo AI Services, you provide your explicit acknowledgment and consent to the data practices defined herein. Specifically, you authorize:
- The automated processing of your submitted information by AI systems.
- The systemic storage of your historical prompts and generated outputs.
- The continuous collection of technical usage analytics.
- The secure transmission and processing of data by validated third-party AI providers.
- The utilization of pseudonymised, de-identified, or masked information to train, refine, and improve Kriyo's proprietary AI models and workflows.
11. Changes to this Policy
As privacy regulations evolve and our AI technologies mature, Kriyo may periodically update this Privacy Policy. When revisions are made, the "Effective Date" at the top of this page will be amended accordingly, and material updates may be accompanied by a prominent system notification. Your continued use of the Kriyo AI Services following the publication of an updated policy constitutes your formal acceptance of the modified terms.