Skip to Content
LegalPrivacy Policy

Privacy Policy

Draft — name frozen as Ticora (vault ADR-7); {{LEGAL_ENTITY}} is still a placeholder pending incorporation. This template has not been reviewed by counsel.

Last updated: 2026-07-19 · Effective: on public launch

This policy describes what the hosted Ticora service (the “Service”), the documentation site, and the marketing site collect — and, just as importantly, what they deliberately do not.

The short version

  • Your application’s row data never rests with us. The SDK fetches your data inside your application; nothing from your fetch results is sent to or stored by the Service.
  • SDK telemetry is off by default, opt-in, and anonymous — a pinned enum of events with no user identifiers, validated server-side; anything outside the schema is rejected.
  • No tracking cookies. The docs and marketing sites set no cookies at all; the dashboard uses only an essential session cookie once session auth ships.

What we store, per surface

Hosted service (tenant data)

To operate your tenant we store: tenant name and settings; workspace specs and their full version history; sharing records; hashed API keys with their scopes and last-used timestamps; and an append-only audit trail. An audit entry contains: tenant id, the acting user id, the action name, a small JSON detail object (e.g. which workspace, which key id — not spec contents), and a timestamp. Audit and version history are retained for the life of the tenant — including through workspace soft-deletes — because they are the tamper-evidence the product promises. Deleting your tenant deletes all of it within 30 days, backups included, except records we must keep by law.

SDK telemetry (opt-in, anonymous)

Exactly as documented in Telemetry: five enum-pinned events (provider.mounted, sandbox.rendered, store.first_save, block.degraded, spec.rejected) plus sdkVersion. The ingest endpoint requires an API key purely as an abuse gate and does not persist the key or the tenant; the store has no tenant or user columns, and only aggregates are readable. No IPs are stored, no user identifiers, no spec contents, prompts, titles, or field names — ever. If telemetry is not explicitly enabled, the code path is inert: no queue, no timers, no network.

Dashboard and account

When session auth ships, we store your login identifier (email), a salted credential hash, and session records; the session cookie is essential-only (no advertising or cross-site tracking use).

Docs and marketing sites

Static content. No cookies, no analytics beacons, no third-party trackers. If we ever add analytics it will be cookieless and aggregate, and this policy will be updated first.

Operational logs

Service request logs (IP, endpoint, status, latency) are kept for security and debugging and rotated within 30 days.

What we never collect

Your end-users’ data; your database contents; spec prompts or LLM conversations from your application; browsing behavior across other sites.

Sharing

We share data only with hosting/infrastructure subprocessors needed to run the Service (listed at {{SUBPROCESSOR_PAGE_URL}} before launch), under data-processing terms. We do not sell personal data. We disclose only when legally compelled, and we notify you unless prohibited.

Your rights

Export your specs and audit trail via the API at any time. Request account deletion or a copy of your personal data at {{CONTACT_EMAIL}}. EU/UK users: legal bases are contract performance (tenant data) and consent (telemetry); you may lodge complaints with your supervisory authority.

Changes and contact

Material changes are announced at least 14 days ahead via the dashboard or email. Controller: {{LEGAL_ENTITY}} · {{CONTACT_EMAIL}}.

Last updated on