← Saykeeper

Privacy Policy

Saykeeper records and processes conversations. That is sensitive, so here we explain simply and fully what data we process, who helps us with it, and where.

1. Who is the controller

The controller is Timur Cimerman, sole trader (SZČO), with place of business at 29. augusta 5682/36A, 058 01 Poprad, Slovakia, Company ID (IČO): 56643454, Tax ID (DIČ): 1130022674, registered in the Trade Register of the District Office Poprad, No. 740-44854 (“Saykeeper”, “we”). Contact for data protection: privacy@saykeeper.com. We have not appointed a data protection officer, so please send everything to this address.

2. What data we process

• Account: email address, password (stored only as a secure hash), name if you enter it, language and settings. If you sign in with Google or LinkedIn (where this option is available), we receive your name, email and profile photo from them (only the basic profile data: openid, email, profile). We do not ask for access to your calendar, mail or contacts.
• Recordings: audio and video that you record in the app or upload to it from a file, and screenshots if you add them to a meeting. The Android app saves a recording on your phone first. It is sent to us for processing when you are signed in and online and your plan and the app settings allow it. Otherwise it stays only on the phone.
• Meeting content: transcripts with timestamps and speaker labels (for example “Speaker 1”, or a name if it is said in the conversation), summaries, tasks, minutes, emails, your questions and the assistant's answers, notes, titles, tags and playlists. We also store the fact that you confirmed the notice to participants before recording, and when.
• Sharing and teams: shared links, email addresses of people you gave access to or who requested it, and the members and invitees of your workspaces and teams.
• Payments: Stripe customer and subscription identifiers, subscription status and period. Card details are processed by Stripe. We do not see or store them.
• Support: what you write to us in the app or by email, together with the platform, app version and device type.
• Integrations you turn on: API keys (we store only their hash), your webhook addresses and a log of tool use via MCP.
• Technical data: sign-in cookies, error logs (account identifier, page, error message, browser and app version, without the IP address), a shortened hash of the IP address for abuse protection, and hosting server logs, which contain the IP address.

We do not collect your location, phone contacts, advertising identifiers or push notification tokens (the Android app shows only local notifications on your phone). The app has no ads and no analytics or tracking tools.

Recordings and transcripts contain personal data of the people you meet. For this data, you (or your company) are the controller, and we process it on your behalf as a processor, only to provide you with the service. It is your responsibility to have a legal basis for recording and to inform people (see recording consent). If you need a data processing agreement under Art. 28 GDPR, write to us.

3. Why and on what legal basis

• Providing the service (account, recording, transcription, summary, email, sharing, teams, support): performance of a contract, Art. 6(1)(b) GDPR.
• Payments, invoices and accounting: performance of a contract and legal obligation, Art. 6(1)(b) and (c) GDPR.
• Security, abuse protection (captcha, rate limiting) and bug fixing: our legitimate interest in a secure and working service, Art. 6(1)(f) GDPR.
• Service emails (password reset, invitations, replies to tickets): performance of a contract. We do not send marketing emails.

Providing your email is necessary to create an account. Without it, we cannot provide the service. We do not use automated decision-making with legal effects or profiling.

4. Artificial intelligence

Transcription is done by ElevenLabs (speech recognition model Scribe v2). Summaries, tasks, minutes, answers to your questions and recognition of speaker names are created by the language model Claude Haiku 5.5, and follow-up emails and messages by Claude Sonnet 5.5, both from Anthropic. We send the text to Anthropic directly through its API. We do not use your recordings or transcripts to train any models, and we do not provide them to anyone for that purpose. Under its terms for the commercial API, Anthropic does not use data for training by default and deletes inputs and outputs within 30 days (content flagged by its safety systems as a policy violation may be kept longer). The transcription provider ElevenLabs may keep audio and transcripts under its own rules. All processors may process data only under a data processing agreement, to provide their service to us.

We do not sell your recordings or transcripts, we do not use them for advertising, and no person reads them, except when you explicitly ask us to (for example to fix a bug), when the law requires it, or when we investigate abuse.

5. Who processes data with us and where

Each processor receives only the data it needs for its task.

Transfers outside the EU. The database, recordings and our servers are in the EU (Ireland). Transcription (ElevenLabs) and artificial intelligence (Anthropic) are processed in the USA. Some processors are American companies (Vercel, Stripe, Resend, Cloudflare, Google) and may reach data from there, for example for support or security. For transfers to the USA, we rely on the European Commission's adequacy decision under the EU-U.S. Data Privacy Framework (decision (EU) 2023/1795) where the processor is certified under it (Vercel, ElevenLabs, Stripe, Resend, Cloudflare, Google), and otherwise on the standard contractual clauses approved by the European Commission (decision (EU) 2021/914), which are part of the data processing agreements (Anthropic, Supabase). We will send you a copy of the safeguards on request.

Third-party services that you choose:

• Live captions in Chrome: while you record in Chrome, the app shows captions using the speech recognition built into the browser. Chrome sends microphone audio to Google servers, and Google processes it under its own rules as a separate controller. In other browsers and in the Android app, captions are created by our server through the same service as transcription, and we store nothing from them. The final meeting transcript is always created only after recording, through our processors.
• Sign-in with Google or LinkedIn: these services process your sign-in under their own rules.
• Webhooks, API and MCP: if you turn them on, we send meeting data to the address or tool that you specify.

6. How long we keep data

• Meetings, recordings, transcripts and outputs: until you delete them or close your account. Deleting a meeting also removes its audio and video.
• Account: for as long as the account exists. After a deletion request, we remove the data within 30 days (see account deletion).
• Error logs: 30 days, then they are deleted automatically.
• On your phone: recordings in the Android app stay on the phone until you delete them or uninstall the app.
• Support tickets: for as long as the account exists, and at most 2 years after resolution.
• Invoices and payment records: for the period required by accounting and tax rules (at most 10 years).
• Backups and server logs: automatic database backups and hosting logs are overwritten on a rolling basis. We use them only for restoring the service and for security.
• At AI processors: under their rules, as described in section 4.

7. Your rights

You have the right of access to your data, rectification, erasure, restriction of processing, data portability, and the right to object to processing based on legitimate interest. Just write to privacy@saykeeper.com. We will reply without undue delay and within one month at the latest. You can also view, edit or delete most data yourself in the app.

If you think we are handling your data unlawfully, you can lodge a complaint with a supervisory authority. In Slovakia, this is the Office for Personal Data Protection of the Slovak Republic, Galvaniho 7/B, 821 04 Bratislava (dataprotection.gov.sk). In the Czech Republic, it is the Úřad pro ochranu osobních údajů (uoou.gov.cz). You can also contact the supervisory authority of the country where you live or work.

Are you in someone's recording? The person who made the recording decides about it, so please contact them first. If you do not know them or they do not respond, write to us and we will help.

8. Security

Data is transferred encrypted (HTTPS). Recordings are kept in private storage, and only you and the people you shared them with can open them. A meeting is private by default. Only you can create a public link, and you can cancel it at any time. Anyone with a public link can see the meeting and can ask the assistant about it, unless you turned this option off.

9. Age

Saykeeper is not intended for people under 16 years of age.

10. Changes

If we change this information in a material way, we will tell you in advance by email or in the app. A detailed list of cookies is on the cookies page.

Last updated: 30 October 2026