In short

• We read your sleep and heart data from Apple Health, if you've connected a wearable. We never write anything back.
• Neo, our AI coach, remembers real, durable things you tell it (like a recurring stressor) — not full transcripts of every conversation.
• If something you say to Neo matches patterns linked to self-harm or suicide, we log that it happened (just the fact and time, never what you said) so we can make sure our safety system is actually working. Neo is not a substitute for professional help — if you're in crisis, contact Lifeline (13 11 14) or emergency services directly.
• We never sell your data, and we never use your health data for advertising.
• You can delete your whole account and its data any time from Settings. If you signed in with Apple, email us and we'll take care of it while we finish a required Apple-specific step.
• We use a small number of trusted services to run highmind, listed in full below — we don't share your data with anyone beyond them.

Full Privacy Policy

Last updated: 3 September 2026

1. Who we are

highmind is operated by Joseph Thomas De Navi (ABN 97 837 742 842), trading as HIGHMIND, a sole trader registered in Victoria, Australia ("highmind," "we," "us").

Contact: [email protected]

2. What we collect

Account information — your email address, and if you use Google or Apple Sign-In, the identifier provided by that service. We never receive your Google or Apple password.

Health and biometric data — with your permission, via Apple HealthKit: sleep duration, heart rate variability (HRV), resting heart rate, intraday heart rate, and step count. Read-only — highmind never writes data back to Apple Health.

City — manually entered by you during setup. This is a typed text field, not derived from your device's location — highmind never requests or accesses GPS/location services.

App activity data — resets you complete, day-state check-ins, nudge and notification interactions, screens viewed, and timestamps of these events.

Conversations with Neo — your messages are processed to generate responses. We extract and store durable, thematic facts from these conversations (e.g. a recurring stressor or preference you've mentioned) to provide continuity across sessions. We do not store full conversation transcripts long-term.

Safety-related data — if a message matches patterns associated with self-harm or suicidal language, we log that this occurred (which detection layer, and when) for safety-auditing. We never store the message content itself in this log. This log is retained for 90 days after account deletion for audit and legal-compliance purposes.