Privacy Policy

Last updated: April 28, 2026

Minutes.AI ("Minutes.AI", "we", "our", or "the Service") is operated by Sense G.K. ("Sense G.K."). This Privacy Policy explains what information we collect when you use the Minutes.AI mobile app, the Minutes.AI web experience, and the supporting backend at sense-ai.world; how we use that information; who we share it with; and the rights you have over it.

By using the Service you confirm that you have read and understood this Policy. If you do not agree, please do not use the Service.

AutoTimecard users: the AutoTimecard apps (iOS and Android) are covered by their own section, which applies instead of sections 1–15: AutoTimecard (English) · AutoTimecard（日本語）

AIFrench users: the AIFrench app (iOS) is covered by its own section, which applies instead of sections 1–15: AIFrench (English) · AIFrench（日本語）

Contents

1. Information we collect
2. How we use information
3. Connected workspace tools (OAuth integrations)
4. Email handling
5. Audio recordings and transcripts
6. AI processing and model training
7. Service providers and sub-processors
8. How we share information
9. Data retention
10. Security
11. International data transfers
12. Your rights and choices
13. Children's privacy
14. Changes to this Policy
15. Contact

1. Information we collect

We collect the categories of information listed below. Each category is processed only for the purposes described in section 2.

1.1 Account information

• Email address (used as your sign-in identifier via Firebase Authentication; for Sign in with Apple, an Apple-relayed email may be stored).
• Display name and avatar, if you set them.
• Subscription / billing status, transaction identifiers, and receipt metadata (handled by Apple App Store / RevenueCat; we receive only the entitlement state, never card numbers).

1.2 Meeting content

• Audio recordings you capture in the Minutes.AI mobile app or upload to the Service.
• Transcripts produced from those recordings, including speaker diarization where supported.
• Generated meeting artifacts: summaries, decision logs, action items, drafted Slack posts, drafted emails, drafted Jira / Linear / Asana / Notion entries, decision memos, dashboards, and other AI-produced documents.
• Any text you type into the Service (for example, custom prompt instructions or transcript edits).

1.3 Tokens for connected workspace tools

When you choose to connect a third-party workspace tool (Slack, Notion, Jira, Linear, or Asana) we receive an OAuth access token from that provider on your behalf. We store that token and the minimum identifying metadata required to use it (for example, your Slack workspace ID and team name, your Jira cloud ID, your Notion workspace ID and selected parent page).

See section 3 for the precise scopes we request and how we use these tokens.

1.4 Dispatch / audit logs

When you publish a generated artifact to a connected tool (for example, posting an AI-drafted summary to a Slack channel) we record an audit entry containing: timestamp, destination identifier (e.g. channel name), success or failure status, the identifier returned by the destination (e.g. Slack message timestamp, Jira issue key, Notion page URL), and a 200-character preview of the body. The full body is also retained inside your meeting workspace so that you can review or re-send it.

1.5 Usage and device information

• Device type, operating system version, app version, locale, and time zone.
• Crash reports and performance traces (Firebase Crashlytics / Performance Monitoring).
• Aggregated usage events such as "recording started", "dispatch tapped", etc., used to understand which features are in use. These events do not contain meeting content.
• IP address (recorded transiently by our servers and CDNs for rate-limiting and abuse prevention).

2. How we use information

We use information to:

• Provide the core Service: capture audio, transcribe it, run AI pipelines that detect tasks and generate artifacts, and let you review and edit those artifacts.
• Synchronise your meeting workspaces, transcripts, and generated artifacts across your devices.
• Carry out an action you explicitly approve, such as posting a drafted message to your Slack workspace or filing a drafted ticket in your Jira project.
• Authenticate you, manage your subscription, and prevent abuse of our infrastructure (rate-limiting, anomaly detection).
• Improve the Service: investigate crashes, debug specific failures, measure feature usage, and develop new features. Where we use meeting content for debugging we do so only on minimal samples needed to reproduce a problem and we do not share that content with third parties for that purpose.
• Comply with applicable law.

We do not sell your personal information, and we do not show third-party advertising.

3. Connected workspace tools (OAuth integrations)

The Service can connect to the third-party tools listed below. Each connection is opt-in and per-account: nothing is sent to a tool until you have completed the OAuth flow for that tool, and even then, only when you tap the relevant action (for example, "Post to Slack").

Tokens are stored encrypted at rest in our managed database (Google Firestore) under a path scoped to your account (/users/{your-uid}/integrations/{provider}). Other Minutes.AI users cannot read your tokens. You can revoke any connection at any time from the Settings screen in the app, or from the third-party tool itself; revocation deletes our copy of the token.

For every dispatch action we record an audit entry as described in section 1.4. The body of the artifact we send is the body that you can view and edit in the app before tapping send.

4. Email handling

When the Service drafts an email on your behalf, the draft is opened in the system email composer on your device (Apple Mail, Gmail mobile, Outlook, or whichever email app you have configured as your default). The actual send happens from your own email account, through your own provider; Minutes.AI does not store your email password and does not send email on your behalf from our servers.

5. Audio recordings and transcripts

Audio you capture is uploaded to our backend and processed by one or more speech-to-text providers (see section 7) to produce a transcript. The audio file and the transcript are then stored in your Firebase Storage / Firestore workspace, scoped to your account.

You can delete a meeting (and the underlying audio + transcript + generated artifacts) from inside the app at any time. See section 9 for retention defaults and section 12 for full-account deletion.

6. AI processing and model training

We send transcripts and prompts to large-language-model providers (currently Google Gemini and OpenAI) to generate summaries, decision logs, drafted messages, drafted tickets, and other artifacts. These providers act as our processors under written contracts and are bound to use the data only to return a response to our request, not to train their general models on your content.

We do not use your meeting content to train Sense AI's own models without your explicit, separate consent. Aggregate signals (for example, "detection of task type X failed Y times this week") used for our internal product analytics do not contain personal or meeting content.

7. Service providers and sub-processors

We use the following sub-processors. Each is bound by a written data-processing agreement.

• Google Cloud / Firebase (Authentication, Firestore, Cloud Storage, Cloud Functions, Crashlytics, Performance Monitoring): primary backend hosting and account database.
• Railway: hosting for the Sense AI sidecar that runs the agentic pipeline and OAuth callbacks.
• Google Gemini API: large-language-model inference for the AI pipeline.
• OpenAI API: large-language-model inference and Whisper-family speech-to-text.
• Google Cloud Speech-to-Text and Microsoft Azure Speech: automatic speech recognition and speaker diarization for some meetings.
• LiveKit: real-time audio transport for live meetings.
• Apple App Store / RevenueCat: subscription billing.
• SendGrid: outbound transactional email such as account-verification messages and support replies. Not used to relay user-drafted content to third parties (see section 4).
• Slack, Notion, Atlassian (Jira), Linear, Asana: only when you explicitly connect them (see section 3). These providers receive only the content you publish to them through Minutes.AI; they do not receive your full meeting workspace.

8. How we share information

We disclose information only:

• To the sub-processors listed in section 7, in the minimum amount required to perform their function.
• To a connected workspace tool when you tap an action that publishes content to it (section 3).
• To comply with a legally binding request from a court or other competent authority.
• In connection with a corporate transaction (merger, acquisition, or asset sale), in which case the acquirer will be required to honour this Policy or to give you an opportunity to opt out before any change in handling.

We do not sell or rent personal information.

9. Data retention

• Account information is retained for as long as your account is active.
• Meeting audio, transcripts, and generated artifacts are retained until you delete them or until you delete your account.
• OAuth tokens are retained until you tap "Disconnect" in Settings, until you revoke the connection from the third-party tool, or until you delete your account.
• Dispatch / audit logs are retained for at least 12 months for fraud prevention and to let you re-send a published artifact.
• Crash reports and aggregated usage events are retained for up to 14 months.
• Backups: residual copies inside our routine backups are overwritten within 35 days.

10. Security

• Data is encrypted in transit (TLS 1.2 or higher) and at rest by default (Google Cloud / Firebase managed encryption).
• Access to production data is restricted to a small number of Sense G.K. engineers, gated by SSO and audited.
• Firestore security rules ensure that a user can only read their own profile and integration records.
• We rotate OAuth client secrets on a regular schedule and as soon as we suspect any compromise.

No system can guarantee perfect security. If you believe you have found a security issue, please contact us at the address in section 15.

11. International data transfers

Our infrastructure is operated primarily in Google Cloud regions in the United States and Asia-Pacific. By using the Service you acknowledge that your data may be processed outside your country of residence, including in jurisdictions whose data-protection laws may differ from those of your home country. Where required we rely on the Standard Contractual Clauses (or equivalent) with our sub-processors.

12. Your rights and choices

Depending on where you live (for example, the EEA, the United Kingdom, California, or Japan), you have some or all of the following rights with respect to your personal information:

• Access a copy of the data we hold about you.
• Correct data that is inaccurate or incomplete.
• Delete your account and the associated data.
• Disconnect any connected workspace tool, which deletes our copy of the OAuth token for that tool.
• Object to or restrict certain processing.
• Export your meeting content and transcripts in a machine-readable format.
• Withdraw any consent you previously gave (where processing is based on consent), without affecting the lawfulness of processing carried out before withdrawal.
• Lodge a complaint with your local data-protection authority.

To exercise any of these rights, use the in-app controls (Settings → Account → Delete account; Settings → Connected tools → Disconnect) or email us at the address in section 15. We respond within 30 days.

13. Children's privacy

The Service is not directed to children under 13 (or the age of digital consent in your jurisdiction, whichever is higher) and we do not knowingly collect personal information from them. If you believe a child has provided us with personal information, please contact us so we can delete it.

14. Changes to this Policy

We may update this Policy from time to time. The "Last updated" date at the top of this page reflects the most recent revision. For material changes (such as new categories of shared data) we will give you advance notice in the app or by email before the change takes effect.

Last updated: April 28, 2026.

15. Contact

For privacy questions, requests to exercise the rights in section 12, or security disclosures, contact:

Sense G.K.
Email: info@sense-ai.world

AutoTimecard

Effective: September 23, 2026

This section applies to the AutoTimecard apps for iOS and Android and to the team relay that supports them, operated by Sense G.K. It replaces the Minutes.AI sections above for AutoTimecard. AutoTimecard records when you arrive at and leave the workplaces you register, and keeps those records on your phone. There is no account and no login.

1. Data that stays on your phone

Everything you enter or the app records is stored on your phone only:

• workplaces (name, map position, radius, time zone),
• employers, work rules and pay rules,
• work records (arrival, departure, breaks, your corrections and their history),
• reports (PDF, CSV, XLSX) until you share them,
• backups you make.

You can view, correct, export and delete all of it in the app. "Delete all data on this device" in Settings deletes every record and setting on the phone. Uninstalling the app deletes its data. Backups you save to Files, iCloud Drive, Google Drive or another location go where you put them and follow that service's terms. Backups never contain team passphrases or plan information.

2. Location

Automatic recording uses the operating system's region monitoring for the workplaces you switched on. The system tells the app when your phone enters or leaves a workplace's radius; the app stores those moments as arrival and departure.

• The app does not record your route, your position between workplaces, places you stop at, or any location outside the workplaces you registered.
• Monitoring stays active outside working hours so that your next arrival is detected. You can turn it off per workplace or entirely at any time; the app then removes its monitoring conditions from the system.
• Location data is never sent off your phone. Not to us, not to a team admin, not to anyone. (Section 5 describes the approximate region Google derives from the connection for usage statistics.)
• Map search and the map view use Apple Maps (iOS) or Google Maps (Android). Your search text and the map area are handled by those services under their own privacy terms.

Automatic recording is an estimate of presence at a workplace. It cannot know whether you were working, on a break, or what you were doing.

3. Teams

A team lets an admin share workplace settings with members and see members' clock-in and clock-out times. Joining a team is an explicit action: you scan the admin's invitation and enter the team passphrase. The join screen states what the admin will see. For teams, the apps use a relay server we operate (hosted on Vercel). The relay stores, per team:

• the team ID and a hash of the team access token (derived from the passphrase; the passphrase itself is never sent),
• each member's roster entry (display name, role, platform, app version), encrypted on the phone with the team passphrase,
• each member's daily attendance summary (date, clock-in and clock-out times and how they were recorded, worked and break durations, current state), encrypted on the phone with the team passphrase,
• the team's shared settings and workplaces, encrypted on the phone with the team passphrase,
• for a team plan: which store it was bought on, the product, the store's transaction ID or purchase token, and the expiry date.

The relay cannot decrypt the encrypted entries: only phones that hold the team passphrase can read them. The relay never receives coordinates, records outside team workplaces, pay amounts, or the records themselves. Attendance summaries are kept on the relay for 120 days and then deleted. When you leave a team, your roster entry and attendance summaries are deleted from the relay. When an admin removes a team, its plan record is released.

The relay keeps standard server logs (request time, path, IP address) for a short period for operation and abuse prevention, under Vercel's log retention. Repeated wrong access attempts against a team are counted to slow down guessing.

4. Purchases

Subscriptions are bought through Apple's App Store or Google Play. We never see your payment details. For a team plan, the app sends the store's signed transaction (Apple) or purchase token (Google) to the relay so the relay can confirm the plan with the store and enable the team's members; the relay keeps the transaction ID or token, product and expiry for that purpose and asks the store again around renewal time. Personal plans are verified on the phone only. If a subscription ends, you keep every record, can still correct and export them, and can still open your backups.

5. Usage statistics and crash reports

To find problems with automatic recording and to learn which parts of the app are used, the app sends usage statistics and crash reports to Google Firebase (Google Analytics for Firebase and Firebase Crashlytics). This is on by default. You can turn it off at any time in Settings with "Share usage data and crash reports"; sending stops immediately on that phone.

What is sent:

• which screens were opened and which features were used (for example "a PDF report was created", "a workplace was saved", "the team board was opened"),
• coarse counts in ranges, such as the number of workplaces or a team's size band (for example "6–10"),
• the state of location permission and automatic recording, error codes from the system's region monitoring, and how long an arrival or departure notice took to reach the app (in ranges such as "1–5 minutes"),
• crash reports (stack trace, device model, OS and app version),
• a random app-instance identifier created by Firebase. Google also derives an approximate region (such as the country) from the connection's IP address.

What is never sent: your phone's location or coordinates, addresses, names of workplaces, employers, teams or people, team IDs or passphrases, record times, pay rates or amounts, anything you type, or advertising identifiers. The data is not linked to your identity and is not used for advertising or to track you across other companies' apps or websites. Google processes this data for us as a service provider under the Firebase terms. Analytics data is kept for 14 months, crash reports for 90 days.

6. What we do not do

• No advertising or tracking SDKs, and no advertising identifiers. The only analytics are the ones described in section 5.
• No account, e-mail address or phone number is required or collected.
• No data is sold or shared with third parties beyond the services named above (Apple, Google including Firebase, Vercel) as needed to run the app.
• No AI service receives your records or settings in this version. If a later version adds an assistant that converts typed settings into rules, it will send only the text you type for that purpose, only when you use it, and this section will be updated first.

7. Children

AutoTimecard is a tool for people who work and is not directed to children under 13 (or the age required in your country).

8. Your choices and rights

Because your data lives on your phone, you exercise your rights in the app: view, correct, export (PDF, CSV, XLSX, backup) and delete. For relay data, leaving a team deletes your entries. Depending on where you live you may have further rights under local law (for example the GDPR, the CCPA or Japan's APPI). For questions about AutoTimecard or to exercise these rights, write to info@sense-ai.world.

9. Changes

We update this section when the apps' behaviour changes and show the effective date at its top.

See also our Terms of Use.