Legal

At Vyteon Ltd ("we," "us," or "our"), we provide elite AI assessment through FluencyIndex.ai. This policy outlines how we handle your data with a "Privacy by Design" approach, ensuring transparency in the Agentic Age.

Account Information: Name, email address, and optional phone number (for OTP verification via Resend). Assessment Content: The text of your prompts and the AI-generated responses within the Gauntlet sandbox. Performance Metrics: Your final 5A Profile scores (Articulation, Audit, Architecture, Adherence, Augmentation). Behavioral Telemetry: Session interaction metadata collected strictly for anti-cheat verification and assessment integrity. The specific signals collected are part of our proprietary verification methodology. Payment Metadata: Payments are handled by Stripe. We do not store card details; we retain only subscription status and transaction IDs.

Contractual Necessity: To provide the assessment and report you have requested. Legitimate Interests: To protect our platform from fraud (telemetry) and to maintain our Global Benchmarking Index. Legal Obligation: To comply with UK tax and financial reporting (HMRC/Stripe).

To provide accurate industry percentiles, Vyteon Ltd maintains a Global Benchmarking Database. Permanent Anonymization: Upon the conclusion of a session or the deletion of an account, your performance scores are moved to our benchmarking engine. De-identification: During this move, all personal identifiers (Name, Email, IP) are permanently stripped. Indefinite Storage: As this anonymized data no longer constitutes "personal data" under UK GDPR, it is retained indefinitely to power our longitudinal industry averages. This ensures the 5A Protocol remains the global standard for AI fluency.

FluencyIndex.ai uses automated systems (Claude 4.5 and Gemini 2.5) to generate competency scores. Logic Involved: Our models evaluate your session against proprietary rubrics across the five 5A dimensions. The specific evaluation criteria and weighting are trade secrets. No Bias: Our algorithms do not use protected characteristics (race, gender, age) as scoring variables. Human Intervention: Under the Data (Use and Access) Act 2025, you have the right to contest an automated decision. If you believe a technical error has affected your score, you may request a Manual Audit by emailing support@vyteon.com. A Vyteon staff member will review your session logs and provide a final determination.

We utilise enterprise-grade APIs to process your data: Anthropic / Google: Your prompts are processed via private API tunnels. Under our enterprise agreements, this data is not used to train their global foundation models. Stripe: For secure subscription management. Resend: For secure OTP and report delivery.

You have the right to access, correct, or delete your personal data. You may also object to our "Legitimate Interest" processing. Note that if you opt out of data contribution for benchmarking, we may restrict your access to the Global Average comparison features of the platform.

We employ HMAC-signed session tokens, AES-256 encryption at rest, and strict RLS (Row Level Security) on our Supabase infrastructure to ensure your assessment remains private and secure.