Flyance
Byungwook Kang ("we", "the operator") processes the personal information of Flyance (the "Service") users in compliance with the Korean Personal Information Protection Act ("PIPA") and related laws. This English policy mirrors the Korean 개인정보 처리방침 and adds, in Section 18, information for users who live in the United States, including California. If the two versions differ, the Korean version governs to the extent Korean law applies; nothing in this English version limits rights you have under the law of your state. This policy is available as "개인정보 처리방침" on the app's sign-in screen, consent screen, and Profile tab.
| Summary | |
|---|---|
| Main information | Two school emails, profile (nickname, gender, birth year, travel taste), trip plans, agent conversations, chat messages and photos, itinerary and expense records, ratings and reports, device token |
| Purposes | School verification and sign-in, matching travel companions, group chat and settlement calculation, user safety, notifications |
| Retention | In general until you leave the Service. Agent conversations 180 days after they close, deleted messages 30 days, notifications 90 days, report and rating records 1 year. If you leave while sanctioned, email hashes used to block re-signup are kept until the restriction ends or for 3 years (Sections 4 and 7) |
| Generative AI | The agent runs on Anthropic's Claude. We never train models on your inputs or outputs (Section 5) |
| Processors and transfers abroad | Supabase (hosting), Anthropic (AI), Apple (push), Resend (verification email). All involve transfers outside Korea (Sections 9 and 10) |
| Disclosure to third parties | None, except where a law requires it or to protect someone from imminent danger to life, body, or property (Section 8) |
| Contact | Privacy officer Byungwook Kang, [email protected] |
We process only the personal information needed for the purposes below. If a purpose changes, we will first ask for separate consent as PIPA Article 18 requires. Retention periods are in Section 4.
The following is necessary to perform our contract with you under the Terms of Service, so we process it without consent under PIPA Article 15(1)(4) (performance of a contract). Where another basis also applies, it is shown in the basis column.
| Purpose | Items | Basis |
|---|---|---|
| Sign-up, school verification, sign-in, preventing duplicate accounts | U.S. school email (.edu, used to sign in), Korean university email (ac.kr), verification times, hash of the verification code, number of attempts, number and times of code sends, your school as determined by the email domain | Art. 15(1)(4) |
| Creating your profile and showing it to other users | Required: nickname, avatar (emoji or image), gender, birth year, travel pace, budget level, interests, smoking, drinking frequency, whether you can drive / Optional: real photo, exchange period, bio | Art. 15(1)(4) |
| Revealing real photos | Per-chat-room consent to reveal your real photo and its withdrawal | Art. 15(1)(4) |
| Registering trips and finding candidates | Cities, dates, date flexibility, desired group size, number of people already going, companion scope, transportation, trip budget and budget level, preferred companion gender, age range and smoking condition, minimum days for partial companionship, trip interests, trip share links | Art. 15(1)(4) |
| Organizing trip conditions and searching through the agent | What you write to the agent, recruiting posts you paste, the agent's replies and result cards, the trip conditions being organized, search conditions and results | Art. 15(1)(4) |
| Agent negotiation and group proposals | Trip conditions used in negotiation, proposed and counter-proposed dates and group sizes, the agent's acceptances, rejections and counter-proposals, notes written by the agent | Art. 15(1)(4) |
| Direct requests and group responses | Requests you send or receive and their messages, your responses to group proposals and votes and when you responded, your group membership status | Art. 15(1)(4) |
| Group chat | Text messages, chat photos, emoji reactions, read positions, records of sharing earlier chat history, summaries for late joiners | Art. 15(1)(4) |
| Report review of deleted chat messages | Text and photo of a message its sender deleted | Art. 15(1)(6) (legitimate interest: other users' safety) |
| Shared itinerary and settlement calculation | Itinerary items (date, time, title, place name, note), expenses (amount, currency, exchange rate, description, date, payer, each person's share), payment records (sender, recipient, amount, confirmation) | Art. 15(1)(4) |
| Trip suggestions based on the academic calendar | Your school, exchange period, suggestions you received and your response, reports of wrong calendar dates | Art. 15(1)(4) |
| Computing and showing altitude (trust score) | Completed trips, number of "would travel again" marks received, response times and missed responses to requests and proposals, dropping out after confirmation, private ratings received, confirmed reports, the computed score | Art. 15(1)(4) |
| Post-trip ratings and notes | Private ratings (stars, tags), "would travel again" marks, short notes left for a companion | Art. 15(1)(4) |
| Receiving and reviewing reports; blocking | Reports (reason, details, reported user, reported message), the operator's decision and notes, your block list, scam-phrase warning records (the message and matched phrase) | Art. 15(1)(4); Art. 15(1)(6) (legitimate interest: user safety) |
| Handling reports and restricting use | Shown on the operator's review screen: the report, the reporting user's nickname, the reported user's nickname, avatar and bio, whether a real photo exists, number of reports received, past sanctions. For a message report, the reported message and the 5 messages before it in the same room (including messages by other users, system messages shown as "시스템" (system), and deleted messages; text and whether a photo is attached) / Sanction records: kind (warning, content hiding, 3-, 7- or 30-day restriction, permanent suspension), reason code, notice shown to the user, period, deciding operator, lift record / Content action records: action (message hidden, bio cleared, real photo cleared), target, related report, acting operator, time | Art. 15(1)(4); Art. 15(1)(6) (legitimate interest: user safety) |
| Receiving and reviewing AI answer reports | Reporting user, IDs of the reported conversation and message, reason (inaccurate information, inappropriate content, personal information exposed, other), optional details written by the user, a copy of the reported AI answer text, status, the operator's decision | Art. 15(1)(4) |
| Preventing sanction evasion (blocking re-signup) | Hashes made with a server secret key (HMAC-SHA256; before hashing, the email is lowercased, trimmed, and any "+tag" in the part before @ is removed) of the U.S. school email and verified Korean university email of an account deleted while restricted or under an unlifted permanent suspension, and the expiry date | Art. 15(1)(6) (legitimate interest: user safety by preventing re-signup to evade a sanction) |
| Push and in-app notifications | Notification settings by type, title and body of notifications you receive (chat notifications include the sender's nickname and up to the first 80 characters of the message), read status. Safety notifications (new-report alerts to operators, the "신고를 검토했어요" (report reviewed) notice to reporters, and, to sanctioned users, the sanction kind and reason notice, "운영자가 콘텐츠를 정리했어요" (content cleared by the operator), "이용 제한이 풀렸어요" (restriction lifted) and "제재 일부가 풀렸어요" (part of a sanction lifted)) cannot be turned off in notification settings. While a user is restricted, only these safety notifications are sent; chat and all other notifications are not. Notification text may be shortened to 300 characters; the full notice is shown in the app | Art. 15(1)(4) |
| Records of agreement to the Terms and Policy and of the AI confirmation | Date and version of the Terms you accepted, date of your AI processing and transfer confirmation | Art. 15(1)(4) |
| Inquiries and disputes | Your email address, the content of your inquiry, how it was resolved | Art. 15(1)(4) |
None. The consent screen item "AI 처리·국외 이전 확인 [필수]" (AI processing and transfer confirmation, required) is an explicit confirmation required by the App Store Review Guidelines. The legal basis under PIPA is performance of the contract (Section 1.1) and the transfer basis in Section 10.
We currently keep no personal information because another law requires it. The Service is free and has no in-app payments.
| Purpose | Items | Basis |
|---|---|---|
| Delivering push notifications | Apple push device token (APNs), token environment, last-seen time | Art. 15(1)(4) |
| Keeping you signed in and preventing misuse | Sign-in times, IP address, device/app information sent by the app (User-Agent) | Art. 15(1)(4) |
| Computing candidates | Your match candidates, candidate counts, records of how counts change by condition | Art. 15(1)(4) |
| Managing AI usage limits and cost | Type and time of each AI call, token counts, cost, result status (no conversation text) | Art. 15(1)(4) |
| Fixing app errors | From the diagnostic reports iOS (Apple MetricKit) gives the app for crashes, hangs, CPU exceptions, and excessive disk writes, only: call stack tree, app version and build, OS version, device type, CPU architecture, exception type, code and signal and termination reason, hang duration, CPU time and disk writes (for those report kinds). Sent over a separate connection that never carries your login (no account token); the server stores no user identifier and records only the date of receipt (not the time). Total intake is capped at 1,000 reports and 30 MB per day, and 128 KB per report. Can be sent before you sign up or confirm. The hosting provider's API gateway may log the IP address of the request | Art. 15(1)(6) (legitimate interest: service stability) |
We keep personal information only as long as needed to perform the contract and delete it as described in Section 6 when the period ends.
| Information | Retention |
|---|---|
| Account and verification data, profile, school, consent and confirmation records | Until you leave the Service |
| Real photo | Until you delete it in the Profile tab or leave the Service. A new photo overwrites the previous one. Deleting the photo also deletes every reveal consent you gave. |
| Consent to reveal your real photo | Until you withdraw it in the chat room, delete your real photo, or leave the Service |
| Trip plans, share links, requests, group responses, block list, academic-calendar suggestions | Until you leave the Service |
| Match candidates | Until you leave the Service (recomputed when conditions change) |
| Records of candidate-count changes | 30 days from creation |
| Agent conversations (including pasted posts and search results) | A conversation closes after 7 days without a new message and is deleted automatically 180 days after it closes; deleted immediately when you leave |
| Agent negotiation records (including participants and the proposals exchanged) | 180 days after the negotiation ends; your participation records are deleted immediately when you leave |
| Chat messages and chat photos, records of sharing earlier chat history, itinerary, expenses, payment records | Kept while the chat room has members; once no member remains in the room, the whole room is deleted (messages, photo files, itinerary, expenses, payment records, summaries). People who left the room do not count, and rooms of finished trips stay while their members remain. Deletion is postponed while a report pointing to the room, its group, or its messages is open or was resolved less than 1 year ago. You can delete your own messages at any time; when you leave the Service, we unlink you as the author and keep them for the remaining members (Section 7) |
| Chat messages you deleted | Text and photo destroyed 30 days after deletion. During those 30 days room members (including the sender) cannot see them; only the operator can see them on the report review screen, when they are the reported message or one of the 5 messages before it. If a report points to the message, it is kept while the report is open and for 1 year after the report is resolved. Previews in the in-app notification list are replaced with "삭제된 메시지예요" (message deleted), but push notifications already delivered to devices cannot be recalled |
| Itinerary items and expenses you deleted (including shares) | Destroyed 30 days after deletion |
| Payment records cancelled or rejected | Destroyed 30 days after cancellation or rejection |
| Late-joiner summaries | Until the recipient leaves or the chat room is deleted; regenerating overwrites it |
| In-app notifications | 90 days from creation |
| Device tokens | Deleted when you sign out, leave, or Apple reports the token is no longer valid. If you turn off notifications in iOS Settings, the app unregisters this device's token the next time you open it |
| Reports and the operator's decisions | 1 year after the report is resolved |
| AI answer reports (including the copy of the reported answer) | 1 year after the report is resolved or dismissed; during that time the copy may outlive the 180-day deletion of agent conversations. If the reporting user leaves, only the link to them is removed |
| Sanction records | Warnings are records that cannot be lifted: 1 year from creation. Time-limited restrictions: 1 year after they end or are lifted. Permanent suspensions: kept while the account exists, or 1 year after being lifted. When a user leaves the Service, all of their sanction records are deleted |
| Content action records | 1 year from creation; records about a user are deleted when that user leaves the Service |
| Email hashes for blocking re-signup | If the account was under a time-limited restriction when deleted, until the restriction ends; if under an unlifted permanent suspension, 3 years after deletion. Destroyed afterwards |
| Messages hidden by the operator | Shown as "운영자가 숨긴 메시지예요" (hidden by the operator) and destroyed under the same rule as messages users delete (30 days after hiding; kept while a report pointing to the message is open and for 1 year after it is resolved). A bio or real photo cleared by the operator is deleted immediately |
| Scam-phrase warning records | Up to 1 year from creation (destroyed together with the message if the message is destroyed first) |
| Private ratings and "would travel again" marks | 1 year after the group ended; altitude already computed is kept |
| Notes left for a companion | 1 year after the group ended; deleted earlier if the recipient leaves |
| Academic-calendar date reports | 1 year from receipt |
| Altitude and its score event records | Until you leave the Service. Score events do not record who rated or marked you, and destroying ratings and marks does not change altitude already computed |
| AI call records (call type, token counts, cost; no conversation text) | 90 days from creation; your identity is removed immediately when you leave |
| Access logs | Sign-in records (sign-in time, IP address, device and app information) are kept for 1 year, then deleted. The hosting provider's API and auth logs are kept for at most 7 days |
| App diagnostic reports | 90 days from receipt. If the request's IP address is logged by the hosting provider's API gateway, that log is kept for at most 7 days |
| Inquiry emails | 1 year after resolution |
The Service's agent exists only to help Korean exchange students in the United States find travel companions. It runs on generative AI (Anthropic, PBC's Claude) and is used for four things:
A feature in which the agent drafts a trip itinerary after a group is confirmed (full delegation) is currently turned off. Before turning it on, we will add its items to this policy and give advance notice under Section 17.
AI-generated replies, negotiation notes, and summaries carry an "AI" badge or an "AI-generated" label in the app. AI output can be wrong, so organized trip conditions are registered only after you confirm them, and a group is confirmed only when you accept it.
| What | Where and what is stored | Basis | Retention |
|---|---|---|---|
| Agent conversations (inputs and replies) | Conversation messages and result cards on our servers (Supabase) | Art. 15(1)(4) | Closes after 7 days without a new message; deleted 180 days after closing; deleted immediately when you leave |
| Negotiation results | Agreed dates, group size, and agent notes in negotiation records | Art. 15(1)(4) | 180 days after the negotiation ends |
| Late-joiner summaries | Saved as the recipient's summary; regenerating overwrites it | Art. 15(1)(4) | Until the recipient leaves or the chat room is deleted |
| Reported AI answers | A copy of the answer text the user reported, saved in the report record | Art. 15(1)(4) | 1 year after the report is resolved or dismissed |
| Raw logs of AI calls | Not stored | Not applicable | Not applicable |
We do not train or fine-tune any AI model on your inputs or outputs, and we do not keep conversations for quality improvement or research.
| Feature | Items sent |
|---|---|
| Agent conversation | The message you just wrote or the post you pasted, the 20 most recent messages in the same conversation (result cards by type only), today's date, the trip conditions being organized (city, dates, date flexibility, desired group size, number of people already going, companion scope, transportation, budget level, preferred companion gender, age range and smoking condition, minimum partial-trip days, interests), and the cities, dates, and status of your other registered trip segments |
| Negotiation: proposing agent | The initiating user's original dates, date flexibility, minimum partial-trip days, desired group size, and travel taste (pace, budget level, interests) / other participants' nicknames and travel taste / the city being negotiated / the date windows in which people can travel together, with their headcount and number of days / dates other agents proposed in the previous round / random identifiers that distinguish trip segments |
| Negotiation: responding agent | The proposed dates and group size, the previous round's response / your original dates, date flexibility, minimum partial-trip days, desired group size, number of people already going, and travel taste |
| Late-joiner summary | The group's city, dates, target size, current headcount, open/closed status, member nicknames, itinerary (date, time, title, place name), expense total and count, totals by payer, who owes whom how much, number of payments awaiting confirmation. Members who have not completed the AI processing and transfer confirmation are shown only as "member". |
Profile email, birth year, gender, real photo, and group chat text are not sent.
You can report any AI answer that has text in the Agent tab (card-only answers cannot be reported; up to 20 reports per person per day), choosing a reason (inaccurate information, inappropriate content, personal information exposed, other) and optionally adding details. The operator is notified, reviews the report in person, and records a status and notes. If needed, the operator can delete the conversation. For review, only the reported AI answer text is copied into the report record; this copy may contain information you typed or about your trip companions. It is destroyed 1 year after the report is resolved or dismissed. If you cannot use the app, email [email protected] with the date and content of the conversation.
The records below were shared with other users and are needed for their settlements, conversation context, and safety. When you leave, we keep their content and remove only the information that identifies you as the author (PIPA Article 15(1)(6), other users' legitimate interests). Other users see these records as from a "deleted user" (탈퇴한 사용자).
| Remaining record | Retention |
|---|---|
| Chat messages, chat photos, records of sharing earlier chat history | Kept while the room has members; once no member remains, the whole room is deleted, including photo files (postponed while a report pointing to the room, its group, or its messages is open or was resolved less than 1 year ago) |
| Itinerary items, expenses and each person's share, payment records | Same as above |
| Private ratings and "would travel again" marks you gave others | 1 year after the group ended |
| Notes you left for a companion | 1 year after the group ended; deleted earlier if the recipient leaves |
| Reports (made by or about you) | 1 year after the report is resolved |
| AI answer reports (including the copy of the reported answer) | 1 year after the report is resolved or dismissed |
| Scam-phrase warning records | Up to 1 year from creation (destroyed together with the message if the message is destroyed first) |
| Academic-calendar date reports | 1 year from receipt |
| Notifications other users received (your nickname and the start of your message in chat notifications) | 90 days from creation |
Separately, if you leave while restricted or under an unlifted permanent suspension, we keep hashes of your two school emails for the period in Section 4 to prevent re-signup that evades the sanction (PIPA Article 15(1)(6)). The hashes are made with a server secret key (HMAC-SHA256; before hashing, the email is lowercased, trimmed, and any "+tag" in the part before @ is removed), and a new sign-up is checked against them only after the person proves they own the address with the email verification code, so no one can learn another person's sanction by typing their email. If the address matches, the new account is restricted until the hash's end date; the person can still sign out or delete the account.
If you wrote your name, contact details, or a photo of your face in these records, please delete those messages yourself before leaving. Deleted messages are destroyed after 30 days as described in Section 4. In-app notification previews are replaced with "message deleted", but push notifications already delivered to other members' devices cannot be recalled.
We do not provide your personal information to third parties, except:
As part of the Service, other users can see:
| Who | What they see |
|---|---|
| All users who completed school verification | Nickname, avatar, travel pace, budget level, interests, smoking/drinking/driving, exchange period, bio, age band (for example "early 20s"), altitude |
| Users who see you as a candidate | The above, plus the city, dates, date flexibility, overlapping dates, number of people already going, and companion scope of that trip segment |
| Users who receive your request | The above, plus your request message |
| Members of the same group | The above, plus group chat, itinerary, and expense and settlement records. Real photos are visible only between members of the same room who have both consented to reveal them. |
| Verified users who open your trip share link | Nickname, avatar, desired group size, number of people already going, open seats, companion scope, transportation, minimum partial-trip days, and each segment's city, dates, and date flexibility |
Your email addresses, birth year, gender, and school names are not shown to other users. While a user is restricted or suspended, they are hidden from other users in candidates, search, and profiles.
We entrust the following work to these providers. Our contracts prohibit processing beyond the assigned work, require security measures, restrict subcontracting, and set liability, and we oversee that they process information safely.
| Provider | Work |
|---|---|
| Supabase Pte. Ltd. | Database operation, sign-in, photo storage, server functions (hosting the Service) |
| Anthropic, PBC | Generative AI processing for the agent (organizing trip conditions, group negotiation, late-joiner summaries) |
| Apple Inc. | Delivering push notifications (Apple Push Notification service) |
| Plus Five Five, Inc. (Resend) | Sending sign-in code emails and Korean university email verification code emails |
Each provider's own sub-processors and their countries are listed on the provider's page:
Exchange rates for settlements come from an external rate service (Frankfurter); we send only currency codes and no personal information.
(1) Hosting: Supabase Pte. Ltd.
| Items | Account and verification data, email hashes for blocking re-signup, profile and real photo, trip plans, agent conversations and negotiation records, requests and group responses, chat messages and photos, itinerary, expense and payment records, ratings, marks and notes, report, block and warning records, notification settings and history, device tokens, consent and confirmation records, access logs, AI call records, AI answer reports, app diagnostic reports without user identifiers |
| Countries | Data storage: United States (Northern Virginia, us-east-1) / contracting entity: Singapore / sub-processors: United States, Singapore, Hungary per Supabase's list of 2026-10-12 |
| When and how | From sign-up, whenever you use the Service, sent over the internet encrypted (TLS) and stored on the provider's servers |
| Recipient and contact | Supabase Pte. Ltd., 65 Chulia Street #38-02/03, OCBC Centre, Singapore 049513, [email protected] |
| Purpose | Storing and processing Service data |
| Retention | The periods in Section 4, or until the service contract ends |
| How to refuse and effect | Do not sign up, or leave with "Delete account" in the Profile tab. All Service data is stored here, so you cannot use the Service without it. |
(2) Generative AI: Anthropic, PBC
| Items | The items in Section 5.3 |
| Countries | United States (storage). Anthropic may route requests for processing to select countries in the United States, Europe, Asia, and Australia. |
| When and how | Each time an agent conversation, automatic negotiation, or summary is processed, sent encrypted (TLS) through the API |
| Recipient and contact | Anthropic, PBC, 548 Market St, PMB 90375, San Francisco, CA 94104, USA, [email protected] |
| Purpose | Organizing trip conditions, negotiating and proposing groups, late-joiner summaries |
| Retention | Deleted within 30 days of receipt or generation. If flagged as violating the Usage Policy, inputs and outputs up to 2 years and safety classification scores up to 7 years |
| How to refuse and effect | Do not complete the confirmation on the consent screen, or after completing it, leave with "Delete account". Agent matching is the core of the Service, so you cannot use it without this transfer. After you leave, nothing further is sent to the AI provider. |
(3) Push notifications: Apple Inc.
| Items | Device token, notification title and body (chat notifications include the sender's nickname and up to the first 80 characters of the message) |
| Country | United States |
| When and how | Each time a notification is created, sent encrypted (TLS) to Apple's push servers |
| Recipient and contact | Apple Inc., One Apple Park Way, Cupertino, CA 95014, USA, https://www.apple.com/legal/privacy/contact/ |
| Purpose | Delivering push notifications to your device |
| Retention | Ends on delivery. If the device is off or offline and cannot receive it right away, Apple keeps it for up to 30 days and delivers it when the device next connects (only the latest notification per app is kept) |
| How to refuse and effect | If you turn off Flyance notifications in iOS Settings, the app unregisters this device's token the next time you open it, and we stop sending to Apple. If you turn off a type in the app's notification settings, we stop sending that type. Either way, the in-app notification list remains available. |
(4) Verification email: Plus Five Five, Inc. (Resend)
| Items | U.S. school email, Korean university email, the email body containing the verification code, sending and delivery records |
| Country | United States (sending region us-east-1). Resend's sub-processors are also all in the United States (list dated 2026-11-05) |
| When and how | Each time you request a code, sent over an encrypted connection |
| Recipient and contact | Plus Five Five, Inc. (Resend), 2261 Market Street #5039, San Francisco, CA 94114, USA, [email protected] |
| Purpose | Sending verification code emails |
| Retention | Email body, addresses and sending records are kept for 30 days from sending, then deleted. Deleted within 90 days after the processing contract ends |
| How to refuse and effect | Do not sign up, or leave with "Delete account". Email verification is required for sign-up and sign-in, so you cannot use the Service without it. |
You can also ask to leave the Service at [email protected].
You may at any time ask to access, correct, or delete your personal information, ask us to stop processing it, withdraw consent, and ask for an explanation or review of an automated decision.
| Right | In the app | By email |
|---|---|---|
| Access | View your profile, trip plans, received requests, block list, and notification settings in the Profile tab | Ask for information not shown in the app |
| Correction | Edit your nickname, avatar, real photo, travel style, exchange period, and bio in the Profile tab | Gender and birth year cannot be changed in the app because they are matching conditions; email us and we will correct them after checking |
| Deletion | Delete your real photo in the Profile tab (this also deletes every reveal consent you gave); delete your own chat messages; delete your account and your information with "Delete account" | Ask us to delete specific information |
| Stop processing / withdraw consent | Withdraw a real-photo reveal in a chat room (from all members of that room; you can reveal again later; not available while restricted, but deleting your real photo withdraws every reveal); turn off push types in notification settings; turn off notifications in iOS Settings; for AI processing see 11.2 | Ask us to stop processing |
| Explanation or review of automated decisions | Not applicable | See Section 12 |
Agent matching is the core of our contract with you, so we cannot provide the Service without AI processing. PIPA Article 37(2)(4) allows us to decline a request to stop processing in this situation unless you clearly ask to end the contract. For this reason we do not offer a way to keep your account while turning off AI processing. To stop AI processing, leave the Service. After that we send nothing further to the AI provider, and your agent conversations are deleted. If you ask by email, we will explain this and confirm whether you want to leave.
Some features run automatically without human involvement. In particular, during negotiation the agent accepts, rejects, or counter-proposes other agents' proposals on your behalf. We consider that these features are not fully automated decisions with a significant effect on your rights or obligations under PIPA Article 37-2(1), for these reasons:
These features also run to perform our contract (Article 15(1)(4)), so under the proviso to that paragraph the right to refuse does not apply. We still publish the criteria and procedures under Article 37-2(4) and accept requests for explanation and review.
| Feature | Who | Information used | Criteria and procedure | Your role |
|---|---|---|---|---|
| Finding candidates | All users who register a trip | City, dates, date flexibility, gender, age, preferred gender and age range, companion scope, smoking and smoking condition, minimum partial-trip days, travel taste, block relationships | Rule-based. Someone is a candidate only if the city matches, dates overlap within the date flexibility, each meets the other's preferred gender and age range, and companion scope and smoking conditions match. Overlaps shorter than the minimum partial-trip length are excluded, and users who blocked each other are excluded. Order: people who both marked "would travel again", then number of overlapping days, then similarity of pace, budget level, and interests, then newest. | You enter the conditions and choose whether to relax them |
| Negotiation: group proposal | Users who completed the AI confirmation and registered a trip | Negotiation items in Section 5.3 | Rules first narrow the possible dates and people; then generative AI proposes dates and a group. The AI does not see other users' gender, age, or private conditions. | The final group is confirmed only when you accept it |
| Negotiation: responding for you | Same as above | Negotiation items in Section 5.3 | If the proposed dates are within your original dates and the group is no larger than your desired size, the proposal is accepted by rule. Otherwise generative AI chooses to accept, counter-propose other dates, or reject, within your date flexibility and minimum days. If the AI call limit is reached or the AI gives no answer, the proposal is rejected. | If a group is proposed as a result, you must accept it within the deadline (48 hours by default, 12 hours if the trip starts within 3 days). If you decline or do not respond, you are not added. |
| Altitude (trust score) | All users | Completed trips, "would travel again" marks received, response times and missed responses to requests and proposals, dropping out after confirmation, private ratings, reports confirmed by the operator | Computed once a day (09:00 UTC) from per-user score events built from the source records; score events do not record who rated or marked you, and destroying the original ratings and marks after 1 year does not change the score. Starts at 3,000 ft, range 0 to 30,000 ft. Ratings and marks are applied over several days so individual timing is hard to infer. Shown on profiles and not used to order candidates. | Reports are confirmed only by a human operator |
| Scam-phrase warnings | Group chat users | Chat message text | Messages are compared with fixed phrase rules (prepayment requests, money-transfer requests, external links), and the recipient sees a warning. Messages are not blocked. | Warnings are informational; reports are reviewed by a person |
| Area | Measures |
|---|---|
| Encryption in transit | Traffic between the app and our servers, and between our servers and providers, is encrypted (HTTPS/TLS) |
| Access control | Row-level security is enabled on every user table with a default of deny. Users read only their own information and what a feature needs. Private ratings and report details cannot be read from the app. |
| Photo protection | Real photos and chat photos are kept in private storage and opened only for people allowed to see them, through links valid for 10 minutes |
| Verification codes | Korean university email codes are stored only as hashes, expire after 10 minutes, and are limited to 5 wrong attempts and 5 resends per hour |
| Minimal administrative access | The server administration key is never included in the app. Only accounts registered as operators can review reports, and an operator cannot handle a report that involves themselves. The operator review screen shows only what is needed to handle a report ("Handling reports and restricting use" in Section 1.1) and is not used for any other purpose. |
| Minimal data sent to AI | Only the items in Section 5.3 are sent, and no raw logs of AI calls are kept |
| Access logs and review | Each time an operator views or handles user information, we record the operator, time, IP address, the report or user handled, and the task, and keep it for 1 year. These records cannot be changed or deleted through the app or the server administration key. Each month the server summarizes the previous month's records, flags anything unusual, and notifies the operator, who reviews them and leaves a record of the review (at least once a month) |
| Privacy officer | Byungwook Kang |
| [email protected] |
Send any question, complaint, or request about personal information to this contact. We will respond and act without delay.
If you cannot resolve a matter with us, or want dispute resolution or advice about a privacy violation, you may contact the following Korean agencies.
| Agency | Phone (in Korea) | Website |
|---|---|---|
| Personal Information Dispute Mediation Committee | 1833-6972 | www.kopico.go.kr |
| Personal Information Infringement Report Center (KISA) | 118 | privacy.kisa.or.kr |
| Korean National Police Agency | 182 | ecrm.police.go.kr |
| Category | Examples from Sections 1 and 2 |
|---|---|
| Identifiers | School email addresses, nickname, account ID, device token, IP address |
| Personal characteristics | Gender, birth year (shown to others only as an age band) |
| User content | Profile photo and bio, trip plans, agent conversations, chat messages and photos, itinerary and expense records, ratings, notes, reports |
| Preferences | Travel pace, budget level, interests, smoking, drinking, driving |
| Service activity | Requests, group responses, notification history, sign-in times, altitude score |
| Diagnostics | Crash, hang, CPU-exception and disk-write reports (call stacks, app and OS versions, device type, CPU architecture, exception details), not linked to your account |
The third parties with whom this information may be shared are the service providers in Section 9 and, as described in Section 8, other users of the Service.
You can review and change most of your information in the Profile tab, and you can delete your account there at any time. For anything you cannot do in the app, email [email protected] (Section 11).
See Section 17. The effective date at the top of this policy shows when it last changed.
The app does not track you across third-party apps or websites, does not use cookies, and does not use advertising or analytics tools. Because there is no cross-site or cross-app tracking, browser "Do Not Track" signals and similar settings do not change how the Service operates.
We do not sell personal information, and we do not share it for cross-context behavioral advertising. We have not done either in the past 12 months. We do not use your information for targeted advertising.
Depending on your state, you may have rights to know, access, correct, or delete your personal information, and to appeal our decision on a request. You can exercise them as described in Section 11. If we deny a request, you may appeal by replying to our decision at [email protected], and we will respond to the appeal in writing.
The Service is not directed to children or to anyone under 18, and we do not knowingly collect information from anyone under 18 (Section 3).