Security Policy

Last updated: 20 October 2026

Our Commitment to Security

At Nurtivelle ("we", "our", or "us", formerly known as TinyTell), security is foundational — not an afterthought. We recognize that pregnancy and wellness data is deeply private, and we have architected our entire infrastructure with that sacred trust in mind.

This document details our technical safeguards, encryption standards, identity management, and vulnerability reporting procedures.

Authentication & Identity

Nurtivelle utilizes Clerk for customer identity and access management:

  • OAuth 2.0 & Social Sign-In: Secure authentication via Google and Apple — we never access or store raw passwords.
  • Email Authentication: One-time secure passcodes and verification codes prevent unauthorized account access.
  • Cryptographic JWT Tokens: Every single API request is signed and verified cryptographically server-side.
  • Session Integrity: Automatic token expiration, rotating refresh tokens, and instant revocation upon logout or account deletion.

Authorization & Strict Data Scoping

We enforce strict boundary enforcement across all backend queries and mutations:

  • Every query and mutation in our cloud backend (Convex) requires an authenticated identity context.
  • Records are strictly scoped to the caller's unique Clerk subject ID. No user can read, query, or mutate another user's personal data.
  • Sensitive endpoints and AI integrations are protected by server-side entitlement checks.

Data Encryption

  • In Transit: 100% of data traveling between mobile devices, backend services, and third-party APIs is encrypted using TLS 1.2 and TLS 1.3 with modern cipher suites.
  • At Rest: All database stores, backups, and underlying volumes in Convex are encrypted at rest using industry-standard AES-256.
  • Device Keychain: Local session tokens on the user's mobile device are protected using Expo SecureStore (Apple Keychain on iOS / Android Keystore on Android).

Rate Limiting & Threat Mitigation

To prevent brute-force attempts, automated abuse, and denial-of-service risks, we maintain multi-layered rate limiting:

  • Per-User Sliding Windows: Mutation endpoints enforce sliding-window quotas to throttle suspicious request velocity.
  • AI Feature Quotas: Resource-intensive features (like nub theory scans) are strictly metered per user account.
  • Automated Blocking: Anomalous traffic patterns trigger immediate defense controls.

Input Validation & Sanitisation

All client submissions are subject to rigorous validation:

  • Schema Validation: Frontend schemas check input shape and sanity before dispatch.
  • Server-Side Type Enforcement: Convex schema validators reject unauthorized fields, oversized payloads, and out-of-bounds parameters.
  • Data Sanitisation: Text inputs in journals and notes are sanitized to eliminate injection and cross-site scripting vulnerabilities.

Stateless Image Processing Protocol

When you use optional ultrasound analysis tools:

  • Images are transmitted over encrypted TLS channels solely for real-time analysis.
  • Analyse-and-Wipe: Once the model produces the result, the image is immediately purged from memory.
  • Images are never written to disk, never placed in long-term storage buckets, and never used to train public machine learning models.

Secret & Key Management

  • Zero Secrets in Client Code: API secret keys, database credentials, and webhook secrets reside exclusively in secured backend environment variables.
  • Public Keys Only on Mobile: Mobile bundles contain only public client tokens necessary for direct SDK handshakes.

Vulnerability Disclosure Program

If you believe you have discovered a vulnerability or security risk within Nurtivelle, we appreciate your cooperation in responsibly disclosing it to us:

  • Email details directly to security@tinytell.app or dale@zylix.ai.
  • Please provide a clear description and reproduction steps where possible.
  • Do not disclose or discuss the vulnerability publicly until our engineering team has reviewed and patched the issue.

We commit to acknowledging receipt of valid security notices within 48 hours and keeping you updated on resolution.

Contact Us

For security inquiries or questions about this policy: