Yellz Privacy Policy (English)
| Item | Details |
|---|---|
| Version | 1.1 |
| Effective date | 2026-10-17 |
| Business responsible | 株式会社NITI Technology |
| Address / representative | 3F Sobunkan Bldg., 6-25-14 Hongo, Bunkyo-ku, Tokyo, Japan; Yuya Kaneko |
| Privacy contact | support@yellz.app |
Yellz is currently not offered in the EU/EEA, the United Kingdom, or Switzerland. Before expanding to those regions, we will add GDPR and equivalent measures to this Policy.
1. Scope and principles
This Policy covers personal information processed through Yellz apps, web pages, the shared live-event database, and support. We do not display advertising or conduct behavioral targeting in the apps. In the apps, we may show commission-based hotel and transport partner links, with the commission disclosed, only for adult users’ upcoming out-of-town shows. We do not use advertising SDKs, and you can turn these links off in settings. Articles on our website (yellz.app) may include commission-based links to partners’ products and services (affiliate links), clearly labeled as ads. We do not load advertising scripts or tracking images for these links, and we do not provide your information to partners. After you open a link, the partner and the affiliate network may use cookies or similar technologies under their own privacy policies to attribute sign-ups or bookings. We do not sell personal information for money or “share” it for cross-context behavioral advertising as defined by California law.[P1]
2. Information we collect
| Category | Examples | Source |
|---|---|---|
| Account | Firebase UID, Apple/Google identifier, email, display name, profile image | User and identity provider |
| Age and locale | Age range or self-declared age, country/region, language, time zone, base currency | OS age signal or user |
| Private live records | Performance, application status/deadlines, seat, notes, expenses, companion links | User input |
| Photos | Attached photos; optional paid cloud backup | User choice; device or cloud processing |
| Contributions/public data | Nickname, performance/setlist proposal, correction, evidence URL, review, report, points, level, badge | User submission and Company calculation |
| OCR/import input | Ticket image, screenshot, full OCR text, forwarded email body, extracted performance/date/venue | User selection or forwarding |
| AI processing | Masked input, structured result, moderation result, model/prompt version, time | Generated during AI use |
| Subscription/payment | Plan, channel, purchase/renewal/expiry, store transaction or Stripe customer/subscription ID, refund state | Apple, Google, Stripe. Card numbers never pass through our servers |
| Usage/device | App actions, searches, settings, device/OS/app version, installation ID, IP, crash/performance data | Firebase tools and server logs |
| Support | Name/contact, message, attachment, response history | User |
| Fan community | Question/answer text, creation time, private authorship links, deletion/moderation state | Submissions and moderation |
| My Live Wrapped | Personal-record aggregates, videos, optional AI counts/comments, generation counts | User generation requests and our calculations |
| Email forwarding/delivery | Dedicated address, receipt counts, drafts, delivery identifiers/status/times | Address issuance and receipt processing |
| Website browsing information | Viewed URLs, referrers, browser and device types, metrics such as page load speed, and estimated country/region (Cloudflare Web Analytics). We do not receive personally identifiable information, including IP addresses | Aggregated by Cloudflare when the website (yellz.app) is viewed. Users are not identified by cookies or local storage |
| Website browsing information (Google Analytics, only with consent) | Viewed URLs, referrers, browser and device types, screen size, language, interactions on the page, estimated country/region, and a random per-device identifier stored in cookies (such as _ga). We do not receive IP addresses | Collected by Google LLC only if you consent to cookies on the website (yellz.app) |
3. Purposes
| Purpose | Main data |
|---|---|
| Authenticate, sync, and present personal records | Account, settings, private records |
| OCR, structure, match, and bulk import | OCR/import and AI processing data |
| Build, verify, and edit shared DB; calculate rankings | Contributions, attendance status, trust score |
| Manage subscriptions, billing, refunds, payment fraud | Payment and account data |
| Send notifications and handle support, reports, appeals | Contact, settings, inquiries, reports |
| Protect security and prevent point manipulation | IP, device, account, payment-method duplicate signals, logs |
| Analyze quality, failures, and aggregate use | Usage/device and crash data |
| Comply with law, protect rights, resolve disputes | Necessary records |
| Display questions/answers and maintain safety | Post text, authorship links, reports/moderation state |
| Create annual summaries/videos and back up/restore photos | Personal aggregates, cloud photos, subscription state |
| Improving the website and understanding page speed | Website browsing information (aggregated) |
3.1 Waitlist and news emails (newsletter)
- The email address, country of residence, optional favorite artist, and language collected by the waitlist form are used to send the confirmation email, to announce the closed beta, and to manage the waitlist. The registration is not finalized, and no announcement is sent, until the email address is confirmed.
- "Send me Yellz news by email (newsletter)" is an optional checkbox that is unchecked by default. You can join the waitlist without checking it, and it is not a condition of using the service.
- Only if you check it, and only after your email address is confirmed, do we use the registered email address and language to send service updates and announcements about new features and events. We record the date and time of your consent and the version of the wording you agreed to.
- You can stop the newsletter at any time using the unsubscribe link in each email (withdrawal of consent). After that we do not send it, and we record the time you unsubscribed. Withdrawal does not affect the lawfulness of use before it. You can ask us to delete the waitlist registration itself at support@yellz.app.
- Adding our official LINE account is optional and is a service provided by LINE. We may show a link to add the account, but we do not link waitlist data to your LINE account information. LINE's own terms and policies govern information after you add the account.
- Emails are sent by our processor Cloudflare, Inc. (Cloudflare Email Service; the U.S. and other disclosed processing regions). Waitlist and newsletter-consent records are kept for 365 days after email confirmation and then deleted.
4. Ticket images, OCR, and email
- OCR ordinarily runs on-device through Google ML Kit.
- We do not store ticket images or full OCR text on Company servers. Where server-side structuring is needed, data is held transiently in process memory. We remove or replace, where practicable, names, phone numbers, email addresses, order numbers, and barcodes/QR codes before sending only necessary portions to AI, then discard temporary data when processing completes.
- Bulk-import screenshots are not persistently stored and are discarded after processing. They are separate from ordinary photos selected for cloud backup.
- For email forwarding registration (F19), Cloudflare Email Workers parses mail forwarded to your dedicated address in memory and sends the subject, body, recipient identifier, and receipt time to our API. Plain text is preferred; HTML-only messages are converted to text. Attachments, sender details, and other headers are not sent to the API. The message limit is 2 MB, and up to 50,000 characters from the subject/body are used for structuring. Attachment-only messages cannot be processed.
- The subject/body are masked before extracting performance details and matching the shared database. Our receiving process does not persist the original message, subject, body, attachments, or sender details; temporary data is discarded when processing ends. Failures do not retain the body or queue the original for automatic reprocessing. Disabled addresses and over-limit deliveries are discarded without extraction.
- Performance details, matches, receipt ID, processing status, and related data are saved as a private draft for you to check and correct before saving a record. Deleting a draft does not delete an already saved record. We also store forwarding addresses, receipt counts, ownership links, and delivery identifiers, status, and times for deduplication. See Section 8 for retention. You can regenerate or disable the address in Settings.
- Separately from our no-persistence design, Cloudflare may temporarily retain delivery processing data and logs under its terms and settings.
- Performance, date, and venue details confirmed and saved by the user become private records or shared-database contributions.
5. AI processing with Vertex AI Gemini
AI submission triage applies to posts in performance threads and artist communities, including questions and answers. Event-day room messages are not assessed by AI or sent to AI. Rule checks reject URLs, images, contact details, prohibited words, and similar violations; subsequent handling uses reports, with automatic hiding when reports exceed a threshold.
| Item | Policy |
|---|---|
| Uses | Structuring OCR/email text, bulk import, moderation, AI search, first-line support |
| Recipient | Google Cloud Vertex AI Gemini (us-central1 or global) |
| Pre-processing | Detect and remove/replace names, phones, emails, order numbers, ticket identifiers; send minimum fields |
| Model training | Google Cloud’s service terms state that Google will not use Customer Data to train or fine-tune AI/ML models without permission or instruction [P2] |
| Retention | We do not store ticket images/full OCR text. Google-side temporary logs, caching, and abuse monitoring may involve temporary retention under Google Cloud’s terms and settings |
| Oversight | Treat AI results as suggestions and combine confirmation, deterministic rules, or human review |
Do not input secrets or another person’s personal information. AI search using Google Search grounding is displayed only to the requesting user and is not persisted to the shared database.
5.1 My Live Wrapped
We aggregate your attended records and the shared catalog into show, venue, city, and confirmed-setlist song counts, top artists, monthly attendance, first attendance date within the year, and busiest month. Seats, expenses, notes, and photos are not read for this calculation. Video summaries and display names are held in memory during rendering.
The short AI comment is off by default in the operator’s configuration. When enabled, the only personal aggregates sent to Vertex AI Gemini are numeric show, venue, and city counts, together with the output language. Names, UID, artist names, venue names, and record text are not sent. A fixed message is used when disabled, pending, or failed. The current implementation has no per-user switch.
Videos are not automatically published; you choose whether to save or share them. They include the above aggregates and information such as artist names, so check the content before sharing. Retention is described in Section 8.
6. Service providers and disclosures
We disclose personal data only with consent, as required by law, in a business transfer, or as needed for the processors and sellers below. Processors act under our instructions. Stores and the MoR may independently process data under their own policies.
| Recipient/provider | Location/processing region | Purpose/data | Role and official information |
|---|---|---|---|
| Google Cloud / Firebase | United States and other disclosed countries; Firestore in the U.S. (us-central1) | Auth, DB, Storage, Cloud Run, Vertex AI, Analytics, Crashlytics, messaging | Processor. Firebase Privacy, Google Cloud DPA [P3][P4] |
| Apple | Store region and other disclosed locations | Sign in with Apple, IAP/refunds, age range; see the end of this section for the refund-review usage disclosure policy | Feature provider/payment party. Apple Privacy |
| Google Play / Google Sign-In | Store region and other disclosed locations | Google sign-in, Play billing/refunds, age signal | Feature provider/payment party. Google Privacy |
| Stripe Managed Payments | U.S. and other Stripe-disclosed locations | Web subscriptions, taxes, refunds, chargebacks, fraud; contact, transaction/device data | MoR/independent seller/payment party. Stripe Privacy, Managed Payments |
| Cloudflare | U.S. and other disclosed regions | Email Workers reception; DNS/CDN/security; email sending (waitlist confirmation and announcement emails via Cloudflare Email Service); event-day room delivery via Durable Objects (planned: messages, identifiers, connection data); Web Analytics for yellz.app (aggregation of page views; see below) | Processor. Cloudflare DPA, Subprocessors |
| Google (Google Analytics) | U.S. and other regions disclosed by Google | Measuring how yellz.app is viewed (only with cookie consent; see below) | Processor. How Google uses data, Google Analytics data protection |
On the website (yellz.app), we use Web Analytics from Cloudflare, Inc. to aggregate how pages are viewed (viewed URLs, referrers, browser and device types, metrics such as page load speed, and estimated country/region). The purposes are to improve the site and to understand page speed. Web Analytics does not identify users with cookies or local storage, and we do not receive personally identifiable information, including IP addresses.
On the website (yellz.app), we use Google Analytics from Google LLC to measure how pages are viewed, only if you consent to cookies. The information sent is the viewed URLs, referrers, browser and device types, screen size, language, interactions on the page, estimated country/region, and a random per-device identifier stored in cookies (such as _ga). The purpose is to improve the site. We do not use Google signals or advertising features, and we do not use this information for advertising or behavioral targeting. If you do not consent, Google Analytics is not loaded. Your choice is stored in your browser's local storage and is not used to identify you. You can withdraw consent at any time from "Cookie settings" at the bottom of the page; withdrawing deletes the Google Analytics cookies. We do not receive IP addresses. For how Google handles this data, see "How Google uses information from sites or apps that use our services". You can also stop measurement with the Google Analytics Opt-out Browser Add-on. The data retention period is 14 months.
Contribution Data may in the future be commercially distributed as B2B data under the Terms. Before doing so, we will remove personal identifiers, add the categories of data/recipient and opt-out method to this Policy, and provide notice or obtain consent where required. There is no current B2B recipient.
For Apple refund reviews, we may provide Apple with usage information related to the purchase in response to Apple’s request (CONSUMPTION_REQUEST). Apple decides refund eligibility.[P11]
When you open a hotel or transport partner link in the app, only the city and the stay dates are sent to the partner. We do not send venue names, locations, artist names, performance IDs, UIDs, email addresses, or advertising IDs. The link opens in an external browser, where the partner’s terms and privacy policy apply. For how the partner handles IP addresses, cookies, and other data sent by your browser, please see the partner’s own explanation.
7. International transfers
Data may be accessed or processed in Japan, the United States, and countries where providers/subprocessors operate. The core database is stored in the Google Cloud Firestore U.S. region us-central1 (Iowa).[P5]
Japan’s APPI requires information/controls for certain transfers to foreign third parties. Korea’s PIPA requires a legal basis and disclosure of transferred items, country, timing/method, recipient, purpose, and retention.[P6][P7]
| Item | Details |
|---|---|
| U.S. data region and transfer method | Google Cloud us-central1 (Iowa); data is transferred continuously over the network as part of providing the Service |
| Legal entity and country of each recipient | Google LLC (U.S.), Stripe, Inc. (U.S.; the Managed Payments seller is shown at checkout), Cloudflare, Inc. (U.S.). For contact details, see each company’s official information in Section 6 |
| Subprocessors and countries | See the subprocessor lists published by each company [P8] |
| Basis for Korean overseas transfer | Disclosed in this Policy as processing outsourced as necessary to perform the contract; where consent is required, we obtain it in advance |
| Information on foreign regimes for Japanese users | See the information on foreign personal information protection systems published by Japan’s Personal Information Protection Commission. Processors implement security measures under their contracts and terms |
8. Retention and deletion
| Data | Retention |
|---|---|
| Account, private records | While active; delete, including backups, within 30 days after account closure |
| Cloud photos | Optional Premium feature, 5 GB (5,000,000,000 bytes) per person. Scheduled for deletion 90 days after expiry. In-app, push, and email notices at expiry and 30 and 7 days before deletion (daily processing may normally delay notices by about one day). On recovery, retain for at least 7 days after creating the final notice. Resubscribing before the deadline avoids deletion; restoration/deletion remain available during retention. Account deletion takes precedence: within 30 days of the request |
| Wrapped/share videos | Expire 24 hours after completion and become eligible for hourly deletion. Signed download URLs last 15 minutes, capped at video expiry. Job records become eligible for deletion after 7 days from creation and after the video is deleted |
| Wrapped AI comments | Processing caches of counts, language, comments, and related data expire 7 days after the request and become eligible for deletion. Ownership links are private. Also deleted on account closure |
| Forwarded-email drafts | Until you delete them or close your account; no automatic expiry is configured. Saved personal records are separate |
| Email forwarding settings/delivery records | Records for known recipients are deleted on account closure. Disabling/regenerating an address does not erase past records. Delivery records, including those for unknown recipients, are deleted 365 days after creation |
| Fan-community questions/answers | Deleting your post clears its text and leaves a deletion marker. Account closure alone retains text and anonymizes authorship links, with no automatic expiry. Deleting a question hides the answer list, including others’ answers; it does not itself erase their text |
| Ticket images, full OCR text, original bulk-import images | No persistent server storage; discard transient data when processing completes |
| Access/security logs | 90 days from collection |
| Shared DB contributions/history | After account closure, unlink and anonymize, then retain indefinitely for database integrity |
| Personal notifications and push delivery records | Deleted immediately when an account closure request is accepted |
| Support inquiries and replies | Remove identifying information, including UID, email, and name, on account closure; retain for 1 year from acceptance of the closure request, then delete. Free-text subjects, messages, replies, AI answers, and attachment links are removed; categories, handling status, and dates remain |
| Purchase/subscription/billing/refund/webhook records | Unlink from the account on closure; retain for 7 years from acceptance of the closure request to support Japanese bookkeeping retention, then delete. Keep only necessary fields, such as transaction identifiers for statutory retention and refunds. Webhook and similar records without an account link are retained for 7 years from processing |
| Reports, rights requests, disputes | 3 years after the matter is closed (or longer where required by law) |
| Firebase-specific data | Follows the retention periods published for each Firebase service [P3] |
Deletion deadlines apply to live data and backups. Records required for compliance, fraud prevention, or disputes are isolated, purpose-limited, and deleted after the required period.
9. Visibility
Public profiles contain only the chosen nickname, level, badges, and contribution count. Attendance history appears only when explicitly made public. Reviewer screens hide the proposer’s name while review is pending. Blocked users do not see one another’s profiles/submissions and are excluded from mutual reviewer selection.
9.1 Fan community
Signed-in users can view free-text questions and answers about artists, including the text and creation time. Author UID, public ID, and nickname are not displayed; the link between author and account is stored privately. Participation does not require a public profile or public attendance history. Minors and users with unknown age categories can only browse. We use rule checks, reporting, staff review, hiding/restoration, and two-way blocking. Question/answer submission checks do not call AI. Posts awaiting moderation or hidden by staff are visible only to their author.
9.2 Public performance/setlist pages (W01) and share links
Pages available without sign-in show verified shared-catalog performance names, dates/times (date only if time is unknown), venue names, artists, cancellation/postponement status, and confirmed setlist titles and order. Pages are open to search engines; search metadata may include the venue’s address. They do not display personal attendance history, seats, notes, expenses, photos, attendance reports, profiles, or nicknames. Performance share links/QR codes include a performance ID and related parameters, but not UID, personal record ID, seats, or notes.
W01 presents shared information: setting a personal record to private or stopping link sharing does not take the page down. There is no user switch to make a performance page private. To request removal from public view, correction, or deletion, contact support@yellz.app with the URL and reason. After staff exclude a page from publication, our delivery cache may retain it for up to one hour. Our settings cannot delete search-engine or third-party copies. Avoid external publication of private images/videos by not sharing them; remove already shared posts at the destination.
9.3 Community expansion and live-room data (planned)
The following features are planned under the owner decisions of 2026-10-09 (#439 and #447). Under-18s and users of unknown age may only browse performance threads, artist communities, and event-day rooms; posting requires age 18 or older and an active account (maintaining #208 eligibility). The adult casual-chat ban is withdrawn: casual-chat categories, event-day rooms, user follows, and reactions are planned for adults. Questions and answers in artist communities undergo rule checks followed by AI submission triage when moderation AI is enabled in the settings. If AI decisions are unavailable because of an outage, cost cap, or emergency shutdown, they are held without publication; while AI is disabled in the settings, they are published after rule checks only.
- We plan to collect/use performance-thread and expanded artist-community text/images (up to 4), room messages/times/authorship links, user/artist follow relationships and approval status, reactions, blocks/muted keywords, report reasons/response history, AI decisions, and trust scores. Purposes are display/delivery, new-content notifications, view preferences, fraud prevention, and safety.
- Age verification continues to prioritize iOS Declared Age Range / Android Play Age Signals, with self-declaration of an age category when unavailable. We do not collect dates of birth. Restricted posting by minors will not be adopted.
- AI moderation sends only necessary content from posts in performance threads and artist communities, including questions and answers, in accordance with the existing AI-recipient disclosures in Sections 5–7 for publication/hold/rejection. AI submission triage applies to posts in performance threads and artist communities, including questions and answers. Event-day room messages are not assessed by AI or sent to AI. Rule checks reject URLs, images, contact details, prohibited words, and similar violations; subsequent handling uses reports, with automatic hiding when reports exceed a threshold. Questions and answers in artist communities undergo rule checks followed by AI submission triage when moderation AI is enabled in the settings. If AI decisions are unavailable because of an outage, cost cap, or emergency shutdown, they are held without publication; while AI is disabled in the settings, they are published after rule checks only.
- Event-day rooms are planned to open 3 hours before doors open at the performance location and close 24 hours after the show ends. They remain read-only for 7 days after closing, then logs are deleted, leaving only non-identifying statistics. Reported content is retained for investigation for 90 days, separated from ordinary logs, accessible only to reviewers, and then deleted. Everyone may browse; posting requires age 18 or older, an active account, and registration of the event as planned/attended. Only text/stamps are allowed, with no URLs/images; rate limits are planned. Existing question/answer retention in Sections 8 and 9.1 does not apply to rooms.
- Event-day room delivery is planned to use Cloudflare Durable Objects. Room messages and identifiers/connection data necessary for delivery are processed by Cloudflare in the United States and other disclosed processing regions. Google Cloud / Firebase remains the authoritative system for authentication, participation registration, and moderation. Sections 6–7 govern processing providers and cross-border transfers.
- Attendance reports are planned to be integrated into the performance thread’s “after the event” tab. Integration alone will not broaden visibility or change existing attendance-report licensing, deletion, or account-closure treatment. Private personal records will not be automatically published, and attendance reports will not appear on W01. Room deletion deadlines do not apply to attendance reports.
9.4 Cheering ads (F26, planned)
Ad-tour check-in location verification (planned) takes place on the device; location data is neither sent to nor stored on servers nor disclosed to other users. Photo location metadata will be removed. Submitted photos, ad placement locations/periods, official announcement URLs, and check-in success results are planned for listing, existence verification, points/badges, and safety review. Ad placement locations are public information distinct from a user’s current location. Rights-holder removal requests use the target, rights concerned, and contact information to handle the request and immediately make the target private.
Terms, Section 4.3 · Community Guidelines, Section 4.3
10. Security
We use reasonable safeguards such as encryption in transit/at rest, access controls, administrator MFA, App Check, personal-data filtering from logs, rate limits, audit logs, and vendor management. No system is absolutely secure. We notify authorities and affected people as required by applicable breach laws.
11. Your rights
After identity verification, you may exercise applicable rights through support@yellz.app. We respond in principle within 30 days (or any shorter deadline required by applicable law).
| Region/law | Main rights | Source |
|---|---|---|
| Japan / APPI | Notice of purpose; disclosure; correction, addition, deletion; cessation/erasure; stopping third-party provision | e-Gov and PPC [P6][P9] |
| California / CCPA/CPRA | Know/access, delete, correct, opt out of sale/sharing, limit sensitive-information use, and non-discrimination. Yellz does not plan sale or behavioral-ad sharing | California DOJ [P1] |
| South Korea / PIPA | Access, correction/deletion, suspension, consent withdrawal, and rights regarding certain automated decisions | National Law Information Center [P7] |
We may request verification and verify an authorized agent. If a statutory exception applies, we will explain the denial. Regardless of whether the CCPA/CPRA applies to us, we provide a comparable request channel within reason.
12. Minors
- The Service is not available under 13 outside Korea or under 14 in Korea.[P10][P7]
- If OS age-range information is unavailable, we request self-declaration.
- Minors’ visibility defaults to private. In the apps, we provide no advertising/behavioral targeting, direct messages, or matching with strangers. We do not show hotel or transport partner links to users under 18 or whose age is unknown.
- If we learn that underage data was collected, we suspend the account and delete it as required. A guardian may contact the privacy channel.
13. Policy changes
Material changes will be notified before their effective date by in-app or other reasonable means. We obtain new consent where required for a new purpose.
14. Contact and regulators
| Purpose | Contact |
|---|---|
| Requests and complaints | support@yellz.app |
| Japan privacy manager | Head of Operations, Yuya Kaneko |
| Korean domestic representative / CPO | Not designated, as we are not currently required to designate a domestic representative. Contact support@yellz.app |
| California request methods | Email to support@yellz.app |
You may also complain to the supervisory authority in your place of residence.
Sources
All sources checked September 24, 2026.
| ID | Primary source | Use |
|---|---|---|
| P1 | California DOJ, CCPA | CCPA/CPRA rights, sale/share, non-discrimination |
| P2 | Google Cloud Service Specific Terms — Generative AI Services | No unauthorized model training/fine-tuning on Customer Data |
| P3 | Firebase Privacy and Security | Service-specific collection and retention |
| P4 | Google Cloud Data Processing Addendum | Processing, deletion, security |
| P5 | Cloud Firestore locations | U.S. multi-regions and location immutability |
| P6 | Japan PPC laws and guidelines | APPI and foreign-transfer guidance |
| P7 | Korean National Law Information Center, PIPA | Rights, under-14 rule, overseas transfer |
| P8 | Google Cloud Subprocessors and Firebase Subprocessors | Subprocessors and countries |
| P9 | e-Gov, Act on the Protection of Personal Information | Japanese rights and third-party disclosure |
| P10 | U.S. FTC, COPPA Rule | Data of children under 13 |
Additional source checked 2026-09-28:
| ID | Primary source | Use |
|---|---|---|
| P11 | Apple ConsumptionRequest | Refund-review usage information and consent requirements |