Tack is an AI meeting driver: it listens to your meetings live, keeps the conversation on track, and turns what was said into decisions, action items, and summaries. That means we handle some of the most sensitive data you have — what you and your colleagues say to each other. This policy explains, plainly, what we collect, where it goes, and what we will and won't do with it.

Tack is currently in an invite-gated closed beta. If anything here is unclear, or you want something corrected or deleted, email us at hong@polytera.dev and we will respond directly.

The short version

We never sell your data, and we don't use your meeting content to train AI models.
Meeting audio is not stored on our servers. It is processed for transcription in real time and discarded.
What we keep is the text: transcripts, speaker names, agendas, decisions, action items, and summaries — stored so you can revisit your meetings.
Audio and transcript text are processed by a small set of specialized service providers (described below) solely to deliver Tack's features.
Public share links are sanitized: they exclude the raw transcript and speaker identities.
You can request deletion of your meetings or your entire account at any time by email.

1. Data we collect

Account information

You sign in with Google. We receive and store your email address and display name. We never see your Google password. We also store the settings you choose in the app (timezone, transcription preferences, default meeting duration).

Meeting content

When you run a meeting with Tack, we store the material the product produces for you: the transcript text, speaker labels and the names you assign to them, agenda topics, detected decisions and action items (including short verbatim transcript excerpts that support them), summaries, and the timeline of events in the session. This content is stored in our database and is scoped to your account and organization.

Meeting audio

Audio from your microphone is captured in your browser and streamed to a transcription provider to be converted into text. We do not keep audio recordings on our servers. A backup recording of the session is held in your browser's memory during the meeting so nothing is lost if the live connection drops; when the meeting ends it is transcribed and then discarded. The one exception: if optional speaker-diarization is enabled, the recording is placed in temporary cloud storage so the diarization provider can read it, and it is deleted immediately after processing — including when processing fails.

Waitlist and referrals

If you join the waitlist, we collect your name and email. If you arrive via a referral link, we record an anonymous visitor identifier and, if you sign up, the association between you and the person who referred you.

Usage and diagnostics

We log metadata about our own AI usage — which feature ran, token and duration counts, and estimated cost — to operate the service responsibly. These logs contain metrics only, never your transcript or prompt text. Our public website uses Google Analytics to understand traffic; see Cookies below.

2. How we use your data

To run your meetings: live transcription, drift detection, topic and decision extraction, and post-meeting summaries.
To let you revisit past meetings and share the outcomes you choose to share.
To gate access during the closed beta (invite codes, allowlist, referrals).
To operate and secure the service: rate limiting, abuse prevention, cost monitoring.

We do not sell or rent your personal data, we do not show ads, and we do not use your meeting content to train our own or anyone else's AI models.

3. Who processes your data

Tack relies on a small set of specialized service providers, each of which receives only what it needs to perform its function on our behalf:

Speech-to-text providers — receive meeting audio to convert it into transcript text.
AI language-model providers — receive transcript text and meeting context to power live moderation, summaries, decisions, and action items.
A speaker-identification provider — receives the session recording to distinguish speakers, only when that optional feature is enabled.
Cloud hosting and database providers — run the application and store your meeting text and account data.
Google — sign-in (OAuth) and website analytics.

These providers process your data solely to provide their service to us, and we select them and configure our usage to keep your content out of model training. All connections are encrypted in transit (TLS), and data in our database is encrypted at rest. We will provide the current list of providers on request at hong@polytera.dev.

4. Sharing and visibility

Your meetings are private to you by default. If you create a public share link, the shared view is deliberately sanitized: it presents the summary, decisions, and action items but excludes the raw transcript, diarized segments, and the real names you assigned to speakers. You can disable a share link at any time.

A small number of Tack administrators can access account and meeting data for support, debugging, and operating the closed beta. We may also disclose data if required by law.

5. Retention and deletion

We retain your meeting content and account data for as long as your account is active, so your meeting history remains available to you. Audio is not retained, as described above.

Self-serve deletion is not yet available in the product. Until it is, email us at hong@polytera.dev from your account email to request deletion of specific meetings or your entire account, and we will complete it within 30 days and confirm.

6. Cookies

We use a session cookie to keep you signed in (it expires after 7 days) and Google Analytics cookies on our public pages to understand website traffic. We do not use advertising or cross-site tracking cookies.

7. Your rights

You may request access to, correction of, or deletion of your personal data, or object to a particular use of it, by emailing us. We honor these requests regardless of where you live, consistent with applicable data-protection laws (including Malaysia's PDPA and, where it applies, the GDPR).

8. Children

Tack is a workplace tool and is not directed at children under 18.

9. Changes to this policy

We will update this policy as the product evolves — for example, when we add self-serve deletion or new integrations. Material changes will be reflected in the effective date above and announced in our changelog.

10. Contact

Questions, concerns, or requests: hong@polytera.dev.