External video disclosure added: September 19, 2026
Article 1 (Purpose)
AboutUs (the “company”) operates VOCO, an AI vocal training service (formerly VOCE, the “service”), in compliance with the Children's Online Privacy Protection Act (COPPA) and other applicable privacy laws. This Privacy Policy explains what information we collect, what we use it for, and how we protect it.
Article 2 (What We Collect and How)
We collect the following information.
When you sign up: email address, password (stored using one-way encryption), and language preference
Information you enter yourself: display name (optional, entered when posting or commenting in the community), post titles and descriptions, conversations with the AI coach, and practice goal settings such as your goal artist; Language, voice, and emotion settings selected for VOCO speech
Audio and video (only when you use those features, after you grant microphone and camera permission):
Voice recordings made during song diagnosis and song practice, and their analysis results
Recordings you upload to the community (Song Showcase) yourself
An audio file and file name you upload, or a source URL you enter, for online target-voice analysis
Still frames from video, sent optionally during diagnosis (to analyze posture and mouth shape, up to 4 images)
Real-time coaching: microphone audio is analyzed on your own device, and only numeric metrics such as pitch and loudness are sent to our servers (the original audio is never sent)
Humming check-in: your device detects only whether a sound occurred, and no recording is saved
Information collected automatically as you use the service: practice records and scores, cookies (Article 10), your IP address (to prevent abuse and enforce rate limits), and service error logs
When you use an in-app purchase: VOCO internal user ID, Apple receipt, product, transaction, subscription, purchase-history and entitlement information, device type, and operating-system information
Push notifications (only if you opt in): push subscription endpoint, the push encryption keys issued by your browser, and your time zone offset (so notifications arrive at your local time)
Information stored only on your device (never sent to our servers): display theme, language, guardian (kids) mode settings, and similar — see Article 10 for details
Article 3 (Why We Process Personal Information)
Identifying members, keeping you signed in, and sending sign-up verification and password reset email
Providing core service features such as song analysis, real-time coaching, and AI diagnosis and coaching results
Storing and displaying practice records, scores, and statistics
Running the community, including displaying posts and handling reports and blocks
Sending practice reminder push notifications (only to users who opted in)
Retrieving in-app products, purchases and restores; validating Apple receipts; managing subscription entitlements; and purchase analytics
Stabilizing the service, including abuse prevention and error diagnosis
Article 4 (Retention and Use Period)
When an account is deleted, account data and files controlled directly by VOCO are deleted without undue delay.
Copies already processed by a service provider are deleted under the applicable contract, provider retention period, and legal obligations. VOCO stops new AI transfers immediately after consent is withdrawn.
Deleting the app or a VOCO account does not cancel an Apple subscription or delete transaction records controlled by Apple. Purchase and transaction records may be retained for the period required by applicable law and provider contracts.
Information we are required by law to retain is kept only for the period that law prescribes.
Article 5 (Processors We Use)
VOCO uses the processors below to provide the service. VOCO limits each processor to the listed purpose and requires appropriate privacy and security obligations under the applicable contract or data-processing terms.
Vercel Inc. — web hosting and privacy-filtered product analytics
Railway Corp. — API hosting, processing and storage of original diagnostic recordings and user files
Neon, Inc. — database hosting for accounts, practice history, settings, and analysis results
Anthropic PBC — Claude API results for AI diagnosis, coaching, visual feedback, comparisons, and personalized plans
Modal Labs, Inc. — online analysis of a target audio file or source URL submitted by the user
Neosapience, Inc. / Typecast US Inc. (Typecast) — server speech synthesis
RevenueCat, Inc. — in-app product retrieval, purchases and restores, Apple receipt validation, subscription entitlements, and purchase analytics
Resend, Inc. or Google LLC (Google Workspace SMTP) — sign-up verification and password-reset email; only the provider configured for the deployment is used
Article 6 (International Transfers)
Because VOCO uses global infrastructure, the information below may be processed outside the user's country. A transfer occurs over an encrypted network when the user requests the relevant feature. The provider and purpose-specific retention rules below apply.
Vercel (United States) — web hosting and privacy-filtered product analytics; retained while the processing service is used and according to applicable logs and legal obligations.
Railway (United States) — API operation and storage of original diagnostic recordings and user files; VOCO-controlled files are deleted without undue delay when the account or applicable data is deleted.
Neon (United States) — account, practice, settings, and analysis database; VOCO-controlled rows are deleted without undue delay when the account or applicable data is deleted.
Anthropic (United States) — receives the data listed in Article 6-2 to generate requested AI results. Under Anthropic's standard API policy, inputs and outputs are ordinarily deleted within 30 days, subject to safety enforcement, legal obligations, or different contractual terms.
Modal Labs (United States) — receives the target audio file, file name, analysis limit, or source URL listed in Article 6-2. Modal states that server-endpoint request and response payloads are not stored; VOCO deletes temporary processing files after the job. Platform logs and operational metadata follow the applicable plan and legal obligations.
Neosapience / Typecast (Republic of Korea or United States, depending on the contracting entity and service region) — receives the speech-synthesis text and settings listed in Article 6-2. Retention follows the enterprise terms applicable to VOCO, Typecast's privacy notice, and legal obligations; VOCO does not state an unverified fixed retention period.
RevenueCat (United States) — receives the internal user ID and Apple receipt, product, transaction, subscription, purchase-history, entitlement, device-type, and operating-system information needed for purchase processing and analytics. Retention and deletion follow VOCO's RevenueCat agreement and applicable legal obligations.
Article 6-2 (Third-Party AI Services and Your Choice)
VOCO sends personal data or user content to the providers below only after a separate in-app disclosure and the user's explicit agreement. General acceptance of the Terms or this Privacy Policy during sign-up is not treated as permission for third-party AI sharing. No listed personal data is sent to these providers before the user agrees.
Anthropic PBC (Claude API)
Purpose: AI diagnosis, coaching, visual feedback, comparisons, and personalized plans
Data sent: Coach messages and recent conversation context; measurements and results derived from voice analysis; voice profile; target artist, practice goals, history and settings; comparison-target details and source URLs; and up to four still frames when the user chooses video analysis. Original diagnostic audio is not sent to Anthropic.
Modal Labs, Inc.
Purpose: Online target-voice analysis using VOCO's analysis pipeline
Data sent: A target-audio file and file name uploaded by the user, or a source URL entered by the user, plus the requested analysis-duration limit.
Neosapience, Inc. / Typecast US Inc. (Typecast)
Purpose: Creating audio that speaks VOCO's reply
Data sent: The reply text to be spoken, language or locale, selected voice ID and model, emotion preset and intensity, and output-format settings. Microphone recordings are not sent to Typecast.
The current production app does not send personal data or user content to OpenAI. Typecast is the only server speech-synthesis recipient. VOCO will update this disclosure, increase the consent version, and obtain a new explicit choice before enabling any different AI recipient.
Original diagnostic recordings may be uploaded to and stored by VOCO's Railway-hosted backend for VOCO's own analysis and account history, but they are not sent to Anthropic. Target audio deliberately submitted for online comparison is sent to Modal only after permission.
If the user does not agree, on-device real-time pitch measurement, basic practice, games, and available on-device speech remain usable. AI diagnosis, coaching, personalized plans, online target-voice analysis, and server speech synthesis are blocked until the user agrees.
The user can review, decline, or withdraw this permission at any time under Settings > Data & Terms > Outside AI Data Sharing. Withdrawal immediately blocks new requests to the three current recipients: Anthropic, Modal, and Typecast. Data already processed by a provider is deleted according to the retention rules in this policy and applicable legal obligations.
VOCO requires each provider, under applicable contracts and data-processing terms, to use the information only for the listed service purpose and to provide the same or an equivalent level of privacy protection, including appropriate security, purpose limitation, and deletion obligations. VOCO will not transmit personal data to a provider if that protection cannot be confirmed.
Article 6-3 (Optional External Teaching Videos)
Some high-note lessons offer an optional YouTube teaching reference. In the mobile app, Watch teacher video on YouTube opens the original video outside VOCO in the YouTube app or your browser; VOCO does not embed the player. On the VOCO website, Open teacher video loads an embedded player only after you choose it. Before either choice, this feature does not load a YouTube player or request a remote video thumbnail. You can continue VOCO practice without opening the video.
Google LLC / YouTube provides the external video service under its own terms and privacy policy. To deliver the requested video and operate its service, Google may receive your IP address, browser and device information, request/referrer information, the selected video and playback activity. Google may process this information outside your country, including in the United States; its policy governs retention and deletion.
For this playback feature, VOCO does not send your VOCO account ID, microphone recordings, practice results or coaching conversation to Google/YouTube. This is viewing an external lesson, not uploading your voice for analysis. Opening the original YouTube link also takes you to an external service governed by Google's policies.
The VOCO website's embedded player uses YouTube's privacy-enhanced mode (youtube-nocookie.com). This does not mean that Google receives no information or that the player is ad-free. YouTube says embedded views in this mode are not used to personalize browsing on YouTube, and ads shown in the player are non-personalized.
Do not open the video or the original YouTube link if you do not want this external connection. On the website, Close video removes the embedded player; it does not erase information already processed by Google. Use Google's privacy controls for that information. VOCO's Outside AI Data Sharing setting covers the AI recipients in Article 6-2 and does not control this separately chosen YouTube service.
Electronic records and files controlled directly by VOCO are deleted using a method designed to prevent recovery when their retention period ends or their purpose is fulfilled.
Deleting a VOCO account does not automatically cancel an Apple subscription or immediately erase records already processed and controlled by Apple or another provider. A user may contact the Privacy Officer to request that VOCO instruct a processor such as RevenueCat to delete eligible customer data. Statutory records and provider security records may remain for the applicable retention period.
Article 8 (Rights of Users and Guardians, and How to Exercise Them)
You may at any time ask to access, correct, delete, or suspend the processing of your personal information. A parent or legal guardian may exercise the same rights over the personal information of a child under 13.
Deleting your account (directly in the app): The Account menu in Settings deletes VOCO-controlled account information, recordings and analysis history, practice and challenge history with daily metrics, AI plans and comparisons, target-artist and goal settings, community posts, comments, likes, reports and blocks, coach conversations, wallet history, settings and onboarding state, push subscriptions, and eligible error logs without undue delay.
Account deletion does not cancel the user's Apple subscription. The user must cancel or manage it in Apple Settings or the App Store. Apple transaction records and records already processed by RevenueCat or an AI provider follow Article 4 and Article 6-2.
Deleting practice data only: From the Data menu in Settings you can delete only your practice history and progress while keeping your account.
Managing posts: You can delete your own community posts and comments yourself, and you can report another user's post or block its author.
For requests that are hard to handle in the app, contact us at the address in Article 12. We will take the necessary steps without delay.
Article 9 (Personal Information of Children Under 13)
A child under 13 needs the consent of a parent or legal guardian to become a member. As a rule, we do not collect personal information from children under 13 without that consent.
The service includes a guardian (kids) mode. With it on, community uploads are blocked and the weekly guardian report is computed only on the user's own device.
Article 10 (Cookies, Automatic Collection, and On-Device Storage)
We use only the cookies that running the service requires:
voce_uid — a session cookie that keeps you signed in and distinguishes users
voce_csrf — a security token cookie that blocks forged requests
You can block cookies in your browser, but some features such as login may stop working. VOCO does not set advertising or tracking cookies for its own service. The optional external YouTube service is subject to the separate processing described in Article 6-3 and Google's policy.
Display theme, language, guardian mode, and some practice statistics are stored only on your device in the browser's local storage (localStorage) and are never sent to our servers. You can erase them at any time by clearing the site data in your browser.
Article 11 (Security Measures)
Passwords are stored with one-way encryption (hashing) that cannot be reversed.
Session cookies are protected from being read by scripts, and traffic is encrypted with HTTPS.
We verify security tokens to prevent forged requests and apply rate limits.
When an account or applicable data is deleted, VOCO deletes the personal data and recordings it controls without undue delay and applies the provider retention and deletion limits described in Articles 4 and 6-2.
Send us any privacy question, complaint, or request for redress and we will respond without delay.
Article 13 (How to Seek Redress)
If you need to report a privacy violation or seek advice, you can contact the following bodies.
U.S. Federal Trade Commission (COPPA enforcement): reportfraud.ftc.gov
FTC Consumer Response Center: 1-877-382-4357
Article 14 (Changes to This Policy)
If this policy is added to, removed from, or amended, we will post a notice in the service starting seven days before it takes effect. If what we collect or why we use it changes materially, we will give notice 30 days in advance.