Loopback 隱私政策 / Privacy Policy
生效日期：2026-07-14 · 最近更新：2026-07-22 · 版本：v1.2
語言與權威版本：本政策提供多語言版本供您閱讀，顯示語言將依您在 Loopback 中選擇的介面語言變化，未提供對應語言時預設回退至英文版。權威版本為英文版；其他語言譯本僅供參考，若譯本與英文版有歧義或衝突，一律以英文版為準。
0. 名詞說明
- 「Loopback」／「我們」：指 HAKKO AI PTE. LTD.（註冊地：新加坡，地址：120 Robinson Road #13-01 Singapore 068913），即 Loopback 反饋分析服務的提供方。
- 「服務」：指 Loopback 提供的多渠道用戶反饋蒐集、彙整與 AI 分析平台（含網頁應用、API 及相關功能）。
- 「客戶」／「您」：指訂閱或使用本服務的企業或組織，及其獲授權之使用者。
- 「終端使用者」：指在客戶所接入的各渠道上留下評論、評分、反饋的第三方個人（如 App Store 評論者、社群成員等）。
- 「個人資訊／個人資料」：可直接或間接識別出特定自然人的資訊。
1. 我們在資料處理中的兩種角色
Loopback 是一款面向企業（B2B）的軟體服務，涉及兩類性質不同的資料，我們的角色也隨之不同：
| 情境 | 資料類型 | 我們的角色 |
|---|
| 客戶註冊、登入、計費、使用平台 | 帳戶與使用數據 | 資料控制者／處理者（Controller）（GDPR）／個人資訊處理者（PIPL） |
| 客戶接入渠道後，平台代其蒐集並分析終端使用者反饋 | 反饋數據 | 資料受託處理者／處理者（Processor）（GDPR）／受託處理者（PIPL），代客戶處理 |
對於反饋數據，處理目的與方式由客戶決定，我們僅依客戶指示處理。相關權利義務另見《資料處理協議（DPA）》（見第 12 節）。
2. 我們蒐集哪些資訊
2.1 帳戶與使用數據（我們作為控制者）
- 帳戶資訊：企業名稱、聯絡人姓名、工作信箱、密碼（加密儲存）、角色/權限。
- 計費資訊：訂閱等級（Free / Pro / Max）、帳單地址、交易紀錄。實際支付卡號由 Stripe 處理，我們不儲存完整卡號。
- 設定資訊：接入的渠道、API 金鑰或授權權杖（加密儲存，詳見第 5 節）。
- 使用與日誌資料：登入紀錄、操作日誌、IP 位址、瀏覽器/裝置資訊、功能使用統計，用於安全、疑難排解與產品改進。
- 產品使用資料：頁面瀏覽與功能使用等產品行為事件（如開啟報告、接入渠道），並附帶一組第一方匿名識別碼（Cookie
lb_aid），用於官網與應用程式之間的匿名歸因。此類資料僅用於產品改進，儲存於我們自有的資料庫（不接入任何第三方分析服務）、不含 Review 內文或您輸入的內容，保留期限不超過 180 天。 - Cookie 及類似技術：詳見第 9 節。
2.2 Review 資料（我們作為受託處理者，代客戶處理）
當客戶接入以下渠道後，平台會依客戶授權，蒐集其中公開或經授權取得的 Review 內容，可能包含個人資訊：
- 應用程式商店：Apple App Store（公開 RSS）、Google Play（Play Developer API）等；
- 社群與工單：Reddit、Discord、Slack、Telegram、Lark / 企業微信、Zendesk、Intercom；
- 調查與口碑：Typeform、Lark 問卷、Trustpilot 等。
這些內容可能包含：Review 內文、評分、發佈時間、平台內顯示的作者暱稱或識別資訊，以及作者主動寫入內容中的其他資訊。我們不會主動索取終端使用者的身分證件、精確定位或其他敏感個人資訊；若 Review 內文偶爾含有此類資訊，我們僅依客戶指示處理，不另作他用。
3. 我們如何使用資訊
- 提供與維護服務：帳戶管理、渠道接入、Review 蒐集與彙整、AI 分類與洞察生成、儀表板呈現。
- AI 分析處理：將 Review 內容送交大型語言模型進行分類、摘要、情感與主題分析（模型清單詳見第 6 節）。
- 計費與客戶支援。
- 安全、防詐欺與法規遵循：異常偵測、存取稽核、履行法律義務。
- 產品改進：基於彙整、去識別化的使用統計優化功能。未經客戶書面同意，我們不會將客戶的 Review 資料用於訓練我們自有的通用模型（第三方模型的訓練政策詳見第 6 節）。
處理的合法性基礎（GDPR / PIPL）
- 履行合約：為提供您已訂閱的服務所必需；
- 正當利益：保障安全、防詐騙、改善產品（已做利益衡平）；
- 同意：適用時（如選用 Cookie、行銷郵件）；
- 法律義務：因應法律法規要求。
- 對於回饋資料，其合法性由客戶作為控制者負責取得與說明。
4. 資料的跨境傳輸與存放
資料儲存與處理位於 Google Cloud（us-central1，美國）。若您位於歐盟/英國，跨境傳輸依據標準合約條款（SCCs）等合法機制進行。
不同部署形態下，資料流向有所不同：
- SHARED_SAAS（共享託管）：資料存於 Loopback 託管環境；
- BYO_KEY（自帶金鑰）：使用您自有的模型/雲端金鑰，AI 呼叫將發送至您指定的供應商；
- SELF_HOSTED（私有部署）：資料存於您自有的環境，Loopback 通常不會接觸您的回饋資料。
5. 資料安全
- 傳輸層採用 TLS 加密；敏感欄位（密碼、第三方 API 金鑰／OAuth 權杖）皆加密儲存。
- 採用以角色為基礎的四級存取控制（RBAC），遵循最小權限原則，並記錄存取稽核。
- 依敏感度提供分層保護（自 L1 資料混淆 至 L4 機密計算，視部署形態與方案而定）。
- 我們採取合理的技術與管理措施，但沒有任何系統能保證絕對安全；發生個人資料外洩時，將依適用法律與合約，於法定時限內通知相關方。
6. 第三方供應商（子處理者 / Subprocessors）
為提供服務，我們會將必要資料委託給以下類別的子處理者。AI 模型的選用取決於您所在的地區與會員等級：
| 類別 | 供應商（範例） | 用途 |
|---|
| 雲端基礎設施 | Google Cloud 等 | 託管、儲存、運算 |
| AI 模型 | Anthropic（Claude）、OpenAI（GPT / Embedding）、Google（Gemini） | 回饋分類、摘要、向量化 |
| 付款 | Stripe 等 | 訂閱計費 |
| 郵件／通知 | Resend | 交易性郵件 |
- 我們與子處理者簽署資料處理條款，要求其提供不低於本政策的保護水準。
- 傳送給模型供應商的內容僅用於產生分析結果；我們優先選用不將客戶資料用於其模型訓練的商用 API 通道，並於合約中明訂此限制。
- 完整、即時的子處理者清單請見：https://loopbackagent.com/subprocessors.html。清單更新時，我們將依 DPA 約定通知客戶。
7. 資料保留
- 帳戶與計費資料：於服務期間及之後，依法律要求（如稅務、審計）保留必要期限。
- 反饋數據：依客戶於平台中的設定或 DPA 約定保留；客戶可請求匯出或刪除。
- 服務終止後，我們將於約定期限（預設 30 天）內刪除或匿名化客戶資料，但法律要求保留者不在此限。
8. 您（及終端使用者）的權利
在適用法律（GDPR／英國 GDPR／加州 CCPA/CPRA／中國 PIPL 等）範圍內，您有權：查閱、更正、刪除、限制處理、資料可攜、反對處理、撤回同意，以及不因行使權利而受到歧視。
- 客戶帳戶資料：請透過第 13 節聯繫我們以行使權利。
- 終端使用者：由於回饋資料由客戶作為控制者掌管，終端使用者的權利請求原則上應向相應客戶提出；我們將在技術上協助客戶回應此類請求。
- 我們將於法定時限內回應；您亦有權向所在地監管機構投訴。
9. Cookie 與類似技術
我們使用必要 Cookie 維持登入與安全，並於取得同意時使用分析類 Cookie 了解產品使用狀況。我們的產品使用分析為第一方自研（頁面不會載入任何第三方分析腳本）；其中第一方匿名識別碼lb_aid僅用於匿名歸因（見第 2.1 節「產品使用資料」），您可透過瀏覽器設定清除或封鎖。您可透過瀏覽器設定管理非必要 Cookie。對於歐洲地區訪客，官網的匿名統計 Cookie（lb_aid）在您於 Cookie 提示中選擇「接受」之前不會寫入，期間亦不收集任何統計事件；選擇「拒絕」則不進行統計，且既有的識別碼會被刪除。無論所在地區，皆可隨時透過頁尾「Cookie 設定」變更選擇。
10. 兒童隱私
本服務面向企業，並非提供給 16 歲以下（或適用法律規定年齡以下）之個人使用。我們不會刻意收集兒童的個人資訊。
11. 政策變更
我們可能不時更新本政策。重大變更將透過服務內通知或郵件告知，並更新頁面頂部「最近更新」日期。變更生效後繼續使用即視為接受。
12. 資料處理協議（DPA）
就我們作為受託處理者處理回饋資料的部分，我們與企業客戶另行簽署《資料處理協議（DPA）》，約定處理範圍、子處理者、安全措施、協助義務、資料外洩通知、審計權及跨境傳輸機制（含 SCCs）。企業客戶可透過第 13 節索取。
13. 聯繫我們
Loopback Privacy Policy
Effective date: 2026-07-14 · Last updated: 2026-07-14 · Version: v1.0
Language & governing version: This Policy is offered in multiple languages for your convenience; the language shown follows the interface language you select in Loopback, and any language not provided defaults to this English version. The governing version is this English version. Other translations are for reference only; in case of ambiguity or conflict between a translation and the English version, the English version prevails.
0. Definitions
- "Loopback" / "we" / "us": HAKKO AI PTE. LTD. (place of registration: Singapore, address: 120 Robinson Road #13-01 Singapore 068913), the provider of the Loopback feedback-analytics service.
- "Service": the Loopback platform for multi-channel collection, aggregation and AI analysis of user feedback (including the web application, API and related features).
- "Customer" / "you": the business or organization that subscribes to or uses the Service, and its authorized users.
- "End User": a third-party individual who leaves comments, ratings or feedback on the channels connected by a Customer (e.g. App Store reviewers, community members).
- "Personal information / personal data": information that can directly or indirectly identify a natural person.
1. Our two roles in data processing
Loopback is a business-to-business (B2B) software service involving two categories of data with different natures, in which our role differs:
| Scenario | Data type | Our role |
|---|
| Customer sign-up, login, billing, platform use | Account & usage data | Data Controller (GDPR) / personal information handler (PIPL) |
| Collecting and analyzing End-User feedback on the Customer's behalf after channels are connected | Feedback data | Data Processor (GDPR) / entrusted party (PIPL), processing on the Customer's instructions |
For Feedback data, the purposes and means of processing are determined by the Customer; we process it only on the Customer's instructions. The respective rights and obligations are further set out in the Data Processing Agreement (DPA) (see Section 12).
2. Information we collect
2.1 Account & usage data (where we are the Controller)
- Account information: company name, contact name, work email, password (stored encrypted), roles/permissions.
- Billing information: subscription tier (Free / Pro / Max), billing address, transaction records. Full card numbers are handled by Stripe; we do not store full card numbers.
- Configuration: connected channels, API keys or authorization tokens (stored encrypted, see Section 5).
- Usage & log data: login records, activity logs, IP address, browser/device information, feature-usage statistics, used for security, troubleshooting and product improvement.
- Cookies and similar technologies: see Section 9.
2.2 Feedback data (where we are the Processor, on the Customer's behalf)
Once a Customer connects the following channels, the platform collects public or authorized feedback content from them as authorized by the Customer, which may contain personal information:
- App stores: Apple App Store (public RSS), Google Play (Play Developer API), etc.;
- Community & ticketing: Reddit, Discord, Slack, Telegram, Feishu / WeCom, Zendesk, Intercom;
- Surveys & reputation: Typeform, Feishu Forms, Trustpilot, etc.
Such content may include: the body of comments/reviews, ratings, publication time, the author's display name or identifier as shown on the platform, and any other information the author voluntarily includes in the content. We do not actively solicit End Users' identity documents, precise location or other sensitive personal information; where feedback content incidentally contains such information, we process it only on the Customer's instructions and for no other purpose.
3. How we use information
- Provide and maintain the Service: account management, channel connection, feedback collection and aggregation, AI classification and insight generation, dashboards.
- AI analysis: sending feedback content to large language models for classification, summarization, sentiment and topic analysis (provider list in Section 6).
- Billing and customer support.
- Security, fraud prevention and compliance: anomaly detection, access auditing, legal obligations.
- Product improvement: optimizing features based on aggregated, de-identified usage statistics. We do not use Customers' Feedback data to train our own general-purpose models without the Customer's written consent (for third-party models' training policies, see Section 6).
Legal bases for processing (GDPR / PIPL)
- Performance of a contract: necessary to provide the subscribed Service;
- Legitimate interests: security, fraud prevention, product improvement (balancing test performed);
- Consent: where applicable (e.g. optional cookies, marketing emails);
- Legal obligation: where required by law.
- For Feedback data, the legal basis is obtained and represented by the Customer as Controller.
4. Cross-border transfers and data residency
Data is stored and processed at Google Cloud (us-central1, United States). If you are in the EU/UK, cross-border transfers rely on lawful mechanisms such as the Standard Contractual Clauses (SCCs).
Data flows differ by deployment form:
- SHARED_SAAS (shared hosting): data is stored in the Loopback-hosted environment;
- BYO_KEY (bring your own key): your own model/cloud keys are used, and AI calls are sent to the provider you designate;
- SELF_HOSTED: data is stored in your own environment, and Loopback generally does not access your Feedback data.
5. Data security
- TLS encryption in transit; sensitive fields (passwords, third-party API keys / OAuth tokens) stored encrypted.
- Role-based access control with four tiers (RBAC), least-privilege principle, access auditing.
- Tiered protection by sensitivity (from L1 data obfuscation to L4 confidential computing, depending on deployment form and plan).
- We apply reasonable technical and organizational measures, but no system can guarantee absolute security. In the event of a personal-data breach, we will notify the relevant parties within the statutory time limits in accordance with applicable law and contract.
6. Third-party providers (Subprocessors)
To provide the Service we entrust necessary data to the following categories of subprocessors. The AI model used depends on your region and membership tier:
| Category | Providers (examples) | Purpose |
|---|
| Cloud infrastructure | Google Cloud etc. | Hosting, storage, compute |
| AI models | Anthropic (Claude), OpenAI (GPT / Embedding), Google (Gemini) | Feedback classification, summarization, vectorization |
| Payments | Stripe etc. | Subscription billing |
| Email/notifications | Resend | Transactional email |
- We enter into data-processing terms with subprocessors requiring protection no lower than this Policy.
- Content sent to model providers is used to generate analysis results; we prioritize commercial API channels that do not use Customer data to train their models, and impose this by contract.
- The complete, up-to-date subprocessor list is available at: https://loopbackagent.com/subprocessors.html. We notify Customers of changes as provided in the DPA.
7. Data retention
- Account & billing data: retained during the service term and thereafter for the period required by law (e.g. tax, audit).
- Feedback data: retained as configured by the Customer in the platform or as agreed in the DPA; Customers may request export or deletion.
- After termination, we will delete or anonymize Customer data within the agreed period (default 30 days), except where retention is required by law.
8. Your (and End Users') rights
To the extent provided by applicable law (GDPR / UK GDPR / California CCPA/CPRA / China PIPL, etc.), you have the right to: access, rectification, erasure, restriction of processing, data portability, objection to processing, and withdrawal of consent, and not to be discriminated against for exercising these rights.
- Customer account data: please contact us via Section 13 to exercise these rights.
- End Users: because Feedback data is controlled by the Customer as Controller, End-User requests should in principle be directed to the relevant Customer; we will provide technical assistance to help Customers respond to such requests.
- We will respond within statutory time limits; you also have the right to lodge a complaint with your local supervisory authority.
9. Cookies and similar technologies
We use necessary cookies to maintain login and security, and, with consent, analytics cookies to understand product usage. You can manage non-essential cookies through your browser settings.
10. Children's privacy
The Service is intended for businesses and not for individuals under 16 (or the age specified by applicable law). We do not knowingly collect children's personal information.
11. Changes to this Policy
We may update this Policy from time to time. Material changes will be communicated via in-service notice or email, and the "Last updated" date at the top will be revised. Continued use after changes take effect constitutes acceptance.
12. Data Processing Agreement (DPA)
For the portion where we act as Processor of Feedback data, we enter into a separate Data Processing Agreement (DPA) with business Customers, covering the scope of processing, subprocessors, security measures, assistance obligations, breach notification, audit rights and cross-border mechanisms (including SCCs). Business Customers may request it via Section 13.
13. Contact us
- Data protection / privacy matters: [email protected]
- General contact: [email protected]
- Mailing address: 120 Robinson Road #13-01 Singapore 068913
- EU/UK Representative (if applicable): to be designated and published separately
- Data protection matters: [email protected]
Loopback