Privacy policy
Mysterious Plant Guide · Effective October 16, 2026 · Operator: NeuralLab / MINSEO LEE
Sending photos and plant names
Photos sent by installed apps with Pl@ntNet consent pass through our app server to Pl@ntNet to find plant candidates. Apps whose new photo-transfer notice names Google Gemini send photos to Gemini after you consent. Existing consent is not treated as consent to send photos to Google Gemini. If identification fails, we do not send the photo to a different provider. The identification provider receives only the photo and a standard identification request, without your app account ID or email. Garden illustrations send only the scientific name and a standard illustration request to OpenAI, without the photo you took. Please avoid photos containing faces or personal information.
Records on your device
Photos, plant names, scientific names, discovery dates and garden layouts are stored on your device. Deleting a plant record also deletes its saved photo. Removing a plant from your garden does not delete its collection record. Uninstalling the app removes its local records. Your operating system’s backup settings may apply separately.
Optional sign-in and account backup
In Collection → Record backup, you can sign in with Apple, Google or Kakao. Supabase and your chosen provider process your unique account ID and any email the provider allows. Google sign-in may also include your name and profile-photo URL in authentication records. We do not request contacts, date of birth, phone number or location. Signing in does not upload guest photos. Importing guest records creates a local account copy. Choosing Back up now sends registered photos, plant names, scientific names, discovery dates, selection status, garden layouts and pots to private Supabase storage. Access is restricted by account. There is no automatic synchronization between devices.
Backup retention and deletion
Deleting a local record does not change your cloud backup until you select Back up now again. A new backup replaces the previous backup, but photo copies retained for failures or restoration may remain until account deletion. One local copy is kept before restoration so you can undo it. Signing out retains that account’s local collection. Deleting your account and backup requires recent sign-in and removes cloud photos, collection, garden, authentication account and that account’s records on this device. Offline copies on other devices and guest records may remain separately. Provider tokens needed to revoke Apple access are stored encrypted on our server and revoked on deletion. Our public account-deletion page and support explain requests without the app. Account and backup data are primarily stored in the Supabase project in Seoul, South Korea. Pro-plan daily database recovery copies may retain pre-deletion data for up to seven days; photo files themselves are not included in these database recovery copies.
App-server processing
Identification photos are handled in memory during requests, without saving them to app-server files or request logs. The server stores shared garden illustrations generated from scientific names and daily request counts. It briefly retains connection IP addresses in memory to limit excessive requests. Railway’s API server processes data in Amsterdam, the Netherlands, and may relay authentication-token exchange and revocation requests. Railway and Supabase process connection information such as IP addresses, request times, status and authentication events for security and troubleshooting. Under the current plans, operators can view the most recent seven days of service logs for each service. Supabase authentication audit logs are not duplicated in a separate database table.
External services
Pl@ntNet’s data-processing policy applies to identification in installed apps. Google’s and OpenAI’s API policies apply respectively to apps with separately enabled Gemini identification and to garden illustrations. Korean-name lookup queries scientific names through sources such as the Korean National Standard Plant List and Wikidata, without sending photos. Processor and transfer details appear below.
Processors and international transfers
We use Railway Corp. (API server in the Netherlands), Supabase (database and private photo storage in Seoul, with operational support from an overseas business), Pl@ntNet/CIRAD (France, identification in installed apps), Google (an overseas business, separately enabled Gemini identification) and OpenAI (a US business, garden illustrations). Photos travel through Railway over HTTPS only to the single identification provider you consented to. Choosing sign-in or backup sends authentication information and your selected backup to Supabase. Each provider’s policies apply to operational support and subprocessors.
Sign-in providers and illustration services
Apple (US), Google (US) and Kakao (South Korea) handle sign-in and disconnection for the account you choose. Garden illustrations send scientific names and a standard illustration request to OpenAI, without app account IDs, email or photos you took. OpenAI API data is not used for model training by default; abuse-monitoring data may generally be retained for up to 30 days. When Gemini identification is offered, it uses a paid API project with request-history storage disabled. Paid inputs and responses are not used to improve Google products or train models, but provider policies apply to limited abuse-prevention retention and legal disclosure. No processing country is fixed; processing may occur in countries where Google or its agents operate facilities. Providers’ legal and safety exceptions apply. This does not mean all retention is absent.
Declining transfers and exercising your rights
If you decline photo transfer, you can view existing records without using identification. You can use local guest records without signing in. Without choosing backup, no account backup of your photos or collection is sent. Contact contact@neuralab.kr to request access, correction, deletion, restriction or withdrawal of consent. We verify identity using minimal information such as your sign-in provider and account email, never passwords or verification codes. We explain the outcome within ten days after identity verification, or explain any restriction or delay and its expected resolution date. A legal representative can also make a request after identity and authority are verified.
Sharing and permissions
Camera access is used to photograph plants; photo selection imports only photos you choose. We do not request location or microphone access. Only choosing Share garden sends an image to your device’s share sheet. The shared image may contain your plant photos. Temporary sharing files are cleaned up on the next share, using a maximum retention period of 24 hours.
Choosing photo identification
When you agree to photo transfer together with the Terms of Service on first launch, the app saves that consent and its time on this device and does not ask again before each identification. You can withdraw consent at any time from the privacy screen, and the app will ask again before the next identification. Without consenting, you can still view your saved collection and garden. You can stop a transfer by cancelling or disconnecting the network, but processing already handed to an external service cannot be undone.
Advertising and your choices
When ads are enabled, the app may show Google AdMob interstitial and banner ads. Interstitials appear infrequently when returning to the camera after checking identification results. A single banner slot appears in Collection or Garden. There are no rewarded ads. Requests are limited to non-personalized ads, with Google’s consent flow first where required. The ad SDK may process IP addresses, device and app information, impressions, clicks and errors for ad delivery, measurement and fraud prevention. We do not pass plant photos, collection contents, account IDs or email to the ad SDK. Non-personalized ads still involve device-data processing. Google and its partners apply their own retention and processing-country policies. Available choices can be changed in the advertising settings in Privacy information. Identification and saving remain usable if ads cannot load.
About the service
Sign-in is optional. We do not use a separate behavioral-analytics SDK. Identification results are suggestions and must not be used to decide edibility, toxicity or medical treatment. Garden illustrations are decorative and may differ from real plants.
Age check and parental permission
On first launch, you enter your birth year and month so the app can work out your age. The year and month are not saved or sent; only the resulting age group and the time a parent or guardian gave permission are stored on your device. Photo identification and record backup are available from age 14 in South Korea and from age 13 elsewhere, and users under 18 need permission from a parent or guardian. Younger users can view only their saved collection and garden. Ad requests include an under-age flag based on the age group.
Processors and transfer details
We entrust providers with authentication, storage and relay operations needed to provide the service. Identification and account backup transfer data only when you choose the corresponding feature. The official notices below explain processing locations and subprocessors.
Pl@ntNet / CIRAD (identification in installed apps)
- Processing location
- France (based on CIRAD’s API address and network registration).
- Data
- Plant photos covered by the installed app’s photo-transfer consent. No app account ID, email or user photo URL is sent.
- When and how
- After Pl@ntNet photo-transfer consent in an installed app, Railway sends the identification request over HTTPS.
- Purpose
- Suggesting plant candidates and managing API usage.
- Retention
- Photos are processed only in volatile memory during identification. API-request history, such as request times, follows the provider’s usage policy, which does not specify a fixed retention period for that history.
Provider policy and contact detailsRailway Corp.
- Processing location
- Amsterdam, the Netherlands (API server). Overseas operational support follows the policy below.
- Data
- Real-time relay of identification photos; connection IP, request time and status; Apple token exchange and revocation requests.
- When and how
- HTTPS during identification requests, Apple connection and deletion.
- Purpose
- API operations, identification relay, account disconnection and security.
- Retention
- Photos remain in memory during requests. Hobby service logs available to the operator cover the most recent 7 days. The provider’s own security records follow its policy.
Provider policy and contact detailsSupabase Inc. / Supabase Pte. Ltd.
- Processing location
- Seoul, South Korea (database and photo storage). See the official list for operational support and subprocessors in the US, Singapore and elsewhere.
- Data
- Account ID, permitted email, Google name/profile URL, authentication information; explicitly backed-up photos, collection, garden and pots; encrypted Apple connection tokens; authentication connection information.
- When and how
- HTTPS when signing in, backing up, restoring or deleting.
- Purpose
- Account authentication, access control and storage of backups you choose.
- Retention
- Accounts and backups remain until account deletion. Old or failed photo copies are also removed on account deletion. Daily database recovery copies cover the most recent 7 days and exclude photo files themselves. Pro logs visible to the operator cover the most recent 7 days. Separate database audit-log storage is disabled.
Provider policy and contact detailsGoogle (Gemini API, when separately enabled)
- Processing location
- Overseas business. No processing country is fixed; processing may occur where Google or its agents operate facilities.
- Data
- Consented plant photo and standard identification request. No app account ID, email or photo URL is sent.
- When and how
- After photo-transfer consent, Railway sends identification requests over HTTPS.
- Purpose
- Suggesting plant candidates and security.
- Retention
- Paid API project. Request-history storage is disabled with generateContent store:false. Paid API inputs and responses are not used for product improvement or model training. Provider policies apply to limited abuse-prevention retention and legal disclosure. There is no separate no-retention agreement.
Provider policy and contact detailsOpenAI
- Processing location
- US business. No separate regional-processing agreement is configured. Official policies govern actual processing locations and subprocessors.
- Data
- Scientific name and standard garden-illustration request. No camera photo, app account ID, email or photo URL is sent.
- When and how
- HTTPS from Railway when requesting a garden illustration.
- Purpose
- Garden-illustration generation and security.
- Retention
- API data is not used for training by default. Abuse-monitoring data may generally be retained for up to 30 days, subject to legal and safety exceptions. No separate Zero Data Retention approval is in place.
Provider policy and contact detailsGoogle AdMob (when ads are enabled)
- Processing location
- Overseas business. Processing may occur in countries where Google and advertising partners operate facilities.
- Data
- IP address, device/app information, ad impressions/clicks and errors. The app does not include photos, collection contents, account IDs or email in ad requests.
- When and how
- HTTPS directly from the device after ads are enabled and any consent process required in the region.
- Purpose
- Non-personalized advertising, measurement and fraud prevention.
- Retention
- Google and advertising partners’ privacy and advertising policies apply. There is no separate no-retention agreement.
Provider policy and contact detailsSupabase subprocessors: Official list · Gemini API terms: https://ai.google.dev/gemini-api/terms · OpenAI data policy: https://developers.openai.com/api/docs/guides/your-data · Supabase contact: privacy@supabase.com. If you do not want a feature that requires international transfers, do not choose photo transfer, sign-in or backup, or contact us below to request a restriction of processing. The corresponding online feature will be unavailable; you can still view records on your device.
Store purchase verification
When you buy a theme or check or restore purchases, Apple or Google handles payment and the Railway app server verifies purchase status through that store’s API. iPhone sends the transaction identifier and Apple-signed transaction data; Android sends the purchase token and product ID. Purchase verification does not receive your app sign-in ID, email, photos or collection records. Our server uses raw purchase data and tokens only in memory while handling the request and does not save them to files, databases or request logs. Store and hosting operational and legal retention policies apply separately. Apple and Google are US businesses, and transactions follow each store’s privacy policy. You can restore purchases with the store account used to buy them. Deleting your app account does not cancel store purchases. Refunded or revoked access is applied when purchase information is next checked. Purchase verification requires an internet connection.
Apple privacy policy · Google privacy policy
Gemini paid API data terms
Gemini request logging
OpenAI API data controls
Supabase privacy
Supabase subprocessors
Railway privacy
Apple privacy
Google advertising data
Google privacy
Kakao privacy
Contact: contact@neuralab.kr