Privacy Policy

Effective August 23, 2026. CTX ("we", "us") builds a service that turns meeting conversations into structured, cited context and routes it into your tools. This page explains what we collect to do that, who else touches it, and what you can do about it.

What we collect

• Account and workspace data — your name, work email, password hash (or nothing at all if you sign in with Google), and the organization you belong to.
• Calendar metadata — meeting times, titles, and participants, read from the Google Calendar you connect, to know which meetings exist and whether our capture policy applies to them.
• Meeting transcripts — spoken content with speaker labels and timestamps, either read directly from a native transcript your meeting platform already produced, or recorded by a bot you explicitly enable as a fallback.
• Generated context — the decisions, requirements, risks, and open questions our extraction produces from a transcript, each one tied back to the exact quote it came from.
• Integration credentials — OAuth tokens for Google and Jira, encrypted at rest, scoped to only what each integration needs.
• Billing data — handled by Stripe; we store your plan and subscription status, never your card number.
• Usage data — API and product usage needed to enforce plan limits and to operate the service reliably.

What we do not do

• We do not sell your data, meeting content, or transcripts to anyone, for any reason.
• We do not use your meeting content to train models for other customers.
• A recording bot never joins a call silently — see Consent below.

How a meeting gets captured

CTX prefers your meeting platform's own native transcript, which requires no bot in the call at all. When that isn't available — the organizer's account doesn't have it enabled, or CTX isn't the organizer — and your workspace has turned on the fallback, a recording bot joins instead. Your workspace's consent setting controls this: it can announce the bot in the meeting chat, require the host to acknowledge it before recording starts, or disable the bot entirely and rely only on native transcripts.

Who else processes this data

We use a small number of subprocessors to run the service, each bound by its own data processing terms:

• Google — calendar and Meet transcript access, under the scopes you grant when connecting Google Workspace.
• Atlassian (Jira) — to create and update tickets on the boards you connect.
• Recall.ai — operates the recording bot, only for meetings where the native tier is unavailable and your workspace enabled the fallback.
• Anthropic — processes transcript text to extract structured context; it does not use API customer data to train its models.
• Stripe — processes payments and stores payment methods; we never see your card details directly.
• Railway and Neon — host the application and its database.

Retention and deletion

Your workspace sets a retention window (30 days by default) after which transcripts and context are deleted on a schedule. Audio from a recording bot, when used, is deleted well before that — within 24 hours of capture. Deleting your workspace deletes everything in it; there is no separate archive we keep around afterward.

Security

• Each workspace's data is isolated at the database level (row-level security), not just by application logic.
• OAuth tokens and other integration secrets are encrypted at rest, not stored as plain text.
• Access to production systems is limited to the people operating the service.

Your rights

You can export or delete your workspace's data, or ask us to do it for you, by writing to the address below. If you're in a jurisdiction with a statutory right to access, correct, or delete personal data (GDPR, CCPA, or similar), we honor that request the same way regardless of where you're writing from.

Changes

We'll update the effective date above when this changes, and post the update here — for anything material, we'll also tell you directly.

Contact

Questions about this policy, or a request about your data: privacy@ctx.dev.