Operator: Beijing Jiefang Silin Network Technology Co., Ltd. ("we", "us").
Privacy contact: serves@silinhome.com
Products covered: /tone/ (tone checkup) and /exam/ (HSK speaking mock exam), including the ToneCoaching Android app, which loads the same web application.
1. What we collect and why
We follow the principles of lawfulness, fairness, necessity and good faith, and we collect only what is needed to make the product work and to keep it secure.
| # | What | Category | How we get it | Why | Required? |
|---|---|---|---|---|---|
| 1 | Your voice recording (the word or sentence you read aloud) | Sensitive personal data — audio can contain voice characteristics | You tap "Record" and grant microphone access in your browser or system prompt | To score pronunciation in real time and draw the pitch curve | Required to use the scoring features |
| 2 | Your native language | Personal data | You choose it from a drop-down before starting | To choose a pronunciation baseline and build a language-specific difficulty profile | Required (an "Other / not listed" option exists) |
| 3 | IP address | Personal data | Automatically, in server logs | Security, rate limiting, abuse prevention, troubleshooting, funnel statistics | Required to operate securely |
| 4 | Browser and device information (User-Agent: browser type/version, OS type/version) | Personal data | Automatically, in web server logs | Compatibility, diagnostics, security auditing | Required to operate securely |
| 5 | Email address | Personal data | You type it into an optional "notify me when the full version launches" form | To email you when the full version launches | Optional — you simply won't get the notice |
| 6 | Student name (max 60 characters) | Personal data | You type it when generating a shareable exam report (may be left blank) | To identify the candidate in a report shared with a teacher or institution | Optional |
| 7 | Institution / teacher name (max 60 characters) | Personal data | You type it when generating a shareable report (may be left blank) | To identify the recipient of the report | Optional |
| 8 | Assessment results: total score, completeness, fluency, per-syllable scores, pitch curve, error list | Personal data (learning record) | Generated automatically by the speech-evaluation provider from your recording | To produce your report, profile and practice list | Required to use the scoring features |
| 9 | Resume marker: session ID, exam level, progress index only | Personal data | Stored in your own browser's localStorage; never uploaded | So you can return to the same mock exam after leaving the page | Optional |
| 10 | Access and operation logs: time, requested URL, response status | Personal data | Automatically | Security auditing and troubleshooting | Required |
| 11 | Usage events, e.g. "answered question 3", "viewed profile", "clicked upgrade" | Personal data (behaviour) | Automatically. Contains neither your audio nor the words you read. | To measure completion and conversion and improve the product | Required |
2. What we do NOT collect
We do not collect or access: your phone number, government ID, facial images, contacts, SMS or call records, precise location, photo library or local files, financial account information. We do not read device identifiers such as IMEI, IDFA or OAID, we do not read your clipboard, and we do not inspect other apps on your device. Apart from the optional fields above, we do not collect your real name, home address or employer.
3. Device permissions
| Permission | When | Why | If you refuse |
|---|---|---|---|
| Microphone | Only when you tap "Record". We never activate the microphone without that action, and we ask for your consent in-app before the first recording. | To record the audio that is scored | Scoring and mock-exam features will not work; browsing still works |
The microphone is the only permission the product requests. We do not request camera, photo library, location, contacts, calendar or notification permissions. Recording happens in your browser, and the page shows a recording indicator while it is active. You can withdraw microphone access at any time in your browser or system settings.
4. Voice recordings and voice characteristics
Your recording may contain voice characteristics that can be treated as sensitive personal data. Before the first recording we show you an in-app notice and ask for your explicit consent. Granting the microphone permission alone is not treated as that consent.
4.1 Our commitments
- Not used to identify you. We do not perform speaker recognition or identity verification, and we do not link recordings to any identity information.
- No voiceprint database. We do not build, retain or share any database of voice characteristics.
- Not used to train our own models. We do not use your recordings to train our own AI models.
- Not kept on our servers. The audio exists only in memory for the duration of that one scoring request. It is released when the request ends and is never written to disk.
- Never sold or traded. We do not sell, rent or otherwise provide your recording to anyone, except where the law requires it.
4.2 Withdrawing this consent
You can withdraw at any time by simply not using the recording features, by revoking microphone permission in your browser or system settings, or by asking us to delete related records (see section 10). Withdrawal does not affect the lawfulness of processing carried out before it.
4.3 Please be careful what you say
While recording, please avoid saying other people's personal or sensitive information (for example ID numbers, bank account numbers or medical details) that is unrelated to the exercise.
5. Third-party processors
To operate the product we use the following processors. Each is bound by a data-processing agreement covering purpose, duration, method, categories of data and security obligations.
| Role | Provider | Data processed | Purpose | Location |
|---|---|---|---|---|
| Speech evaluation | Tencent Cloud Smart Oral Evaluation (SOE) — Tencent Cloud Computing (Beijing) Co., Ltd. | Your voice recording; the reference text of the exercise | Real-time pronunciation scoring and per-syllable analysis | Mainland China |
| Speech synthesis | iFlytek Co., Ltd. (open platform) | Standard reference texts from our question bank (no personal data of yours). Your recording is included only if the backup scoring channel is used. | Generating the model-pronunciation audio; backup scoring channel | Mainland China |
| Cloud hosting and storage | Vultr (The Constant Company, LLC) | The stored data listed in section 1 (IP address, native language, optional email, optional names, learning data, access logs) | Data storage and computation | Singapore |
| SMS / push / crash analytics | Not used in this version. The product has no account system, sends no SMS, and does not integrate any third-party analytics, advertising or crash-reporting SDK. | — | ||
6. How we use your data
- To provide, maintain and improve the core features (tone checkup, speaking mock exam, pronunciation profile and practice list);
- To keep the service and the product secure, and to detect and handle anomalies, fraud and abuse;
- To measure completion and improve product design and teaching effectiveness;
- To respond to your questions, complaints and requests;
- To comply with legal obligations.
We do not use your personal data for any purpose not described in this policy. If we ever want to, we will tell you first and ask for your consent. We do not use your recordings or learning data to train our own AI models. The product does not use personalised recommendation algorithms, does not build advertising profiles, and shows no third-party advertising.
7. Sharing, transfer and public disclosure
We do not share your personal data with third parties except: (a) with your separate consent; (b) where required by law or by a lawful request from a judicial or administrative authority; (c) where necessary to protect someone's life, health or property and consent cannot reasonably be obtained; or (d) as otherwise permitted by law. We do not sell your personal data. We do not transfer it to any company or organisation except in the case of a merger, split, dissolution or bankruptcy, in which case we will tell you the recipient's identity and require them to honour this policy.
7.1 Shared exam links — please read
When you generate a shareable exam report, that link is public. Anyone who has the link can open the report — there is no login and no password. The report contains the student name and institution/teacher name you entered, together with the scores, per-syllable results and pitch curve.
- Prefer a nickname or surname in the "student name" field rather than a full legal name.
- Do not post the link in public forums, group chats or social media.
- If you no longer need the link, you can ask us to delete the stored report snapshot (section 10).
Once a link has been generated, we cannot recall copies that have already been forwarded. Please judge carefully before sharing.
8. Where we store data, and for how long
8.1 Location
The service runs on cloud servers located in Singapore (Vultr). The data listed in section 1 — IP address, native language, optional email and names, learning data, access logs — is stored there. To perform scoring, your recording is transmitted in real time from Singapore to the Tencent Cloud SOE endpoint in mainland China; it is not written to disk (see 4.1).
8.2 International transfers
If you use the product from outside Singapore, your data will be transferred to and processed in Singapore and, for scoring only, in mainland China. Where the law applicable to you requires additional safeguards for such transfers, we rely on your explicit consent for the processing of your voice, and on our legitimate interest in operating a secure service for the technical logs, and we will put appropriate contractual safeguards in place with our processors on request.
If you are located in mainland China, storing your data in Singapore constitutes a transfer of personal data out of China; we handle this in accordance with applicable Chinese law and ask for your separate consent where required.
8.3 Retention
| Data | Retention | What happens after |
|---|---|---|
| Voice recording | Not stored at all. Memory only for the duration of one request; never written to disk. | Not applicable |
| Per-session results (scores, per-syllable data, pitch curve) | Server memory only: 3 hours for a tone checkup session, 6 hours for a mock-exam session; also cleared when the service restarts | Automatically released |
| Usage event logs (IP, native language, optional email, upgrade interest) | No automatic expiry yet. We are deploying automatic cleanup; once live, entries will be deleted 6 months after collection and this policy will be updated. | Deleted or anonymised |
| Shared exam report snapshots (scores, per-syllable data, optional names) | No automatic expiry yet — retained until you ask us to delete them. A link expiring does not mean the snapshot was deleted. | Deleted |
| Server access logs (IP, time, User-Agent) | 14 days (daily rotation, 14 files kept) | Automatically deleted |
Resume marker in browser localStorage | On your own device until you clear site data | You clear it |
9. How we protect your data
Technical: HTTPS/TLS across the site; backend services listen only on loopback addresses and are not directly exposed; credentials are injected via environment variables rather than committed to source; access control and least privilege; regular backups.
Organisational: internal data-security rules, role-based access to personal data, confidentiality obligations for staff, a security-incident response plan, a designated privacy lead, and periodic compliance checks.
If a personal-data breach occurs or is likely, we will activate our incident plan, take remedial measures, and notify you and the competent authority as required by applicable law, including the nature of the incident, the measures taken, what you can do to reduce risk, and any remedies.
10. Your rights
Depending on where you live, you may have the right to access, correct, delete, restrict or object to processing, to withdraw consent, to data portability, and to lodge a complaint with a supervisory authority. We honour these requests without unreasonable conditions.
10.1 Because there is no account
There is no registration and no login, so there is no "delete my account" step. You control your data like this:
- Closing the session: session data in server memory expires after 3 hours (checkup) or 6 hours (exam).
- Clearing browser data: removes the resume marker stored on your device.
- Revoking the microphone: in your browser or system settings, which stops all recording.
- Deleting server-side records: email us and we will remove the event logs and report snapshots associated with you.
10.2 How to exercise them
| Right | How | Our response time |
|---|---|---|
| Access / copy | Your report and profile pages are themselves viewable and saveable (screenshot or print). For server-side records, email serves@silinhome.com with your session ID (shown on the report page), share-link ID, or approximate time of use. | Within 15 working days |
| Rectify | Scores are generated automatically by the evaluation engine and cannot be edited by hand; you can simply retake the test. Optional names or emails can be corrected or deleted by email. | Within 15 working days |
| Erasure | Recordings are not stored, so there is nothing to erase. Clear browser data yourself for the local marker. Email us to delete server-side logs and report snapshots. | Within 15 working days |
| Withdraw consent | Revoke the microphone permission, stop using the recording features, or email us to delete records already created. | Within 15 working days |
| Explanation | Ask us how a policy clause or a result was produced. | Within 15 working days |
| Portability | Because we hold no account and no audio, the transferable content is primarily your report data. Email us and, where the applicable law provides for it, we will provide a route to transfer it. | Within 15 working days |
10.3 Teachers and institutions
If you generate a shared report for a student, please confirm you have the student's consent (and, where the student is a minor, their guardian's consent) and prefer a nickname in the name field. If you entered someone's name without consent, contact us and we will delete it.
11. Children and young learners
This is a pronunciation practice tool, and some users may be school students. The product is intended for learners aged 13 and over, and we show an age gate before use.
- We do not knowingly provide the service to children under 13 (or under 16 in jurisdictions where that is the applicable age of digital consent) without verifiable parental consent.
- Under Chinese law we do not provide the service to children under 14 without a guardian's consent; if we learn we have collected such data without consent, we will delete it promptly.
- Guardians: the core feature requires recording, which may involve a minor's voice characteristics. We recommend that a guardian operates the app with the child, and that the child avoids saying their full name, school or address while recording.
12. Changes to this policy
We may revise this policy from time to time. We will not reduce your rights under it without your explicit consent. Updated versions are published in the product with a new effective date; for material changes we will notify you prominently and, where required, ask for your consent again. Material changes include a significant change in how we process data, a change of ownership, a change in the main recipients of your data (including a new evaluation provider or a change of server country), a significant change in your rights, or a high-risk finding in a data-protection impact assessment.
13. Contact us
Questions, requests or complaints about this policy or your data:
- Email: serves@silinhome.com
- Operator: Beijing Jiefang Silin Network Technology Co., Ltd., Room 519, Hongfu Building, Beiqijia, Changping District, Beijing, China
We respond within 15 working days. If you are not satisfied with our answer, you may also complain to the competent data-protection authority in your country of residence.
14. Relationship to the Terms of Use
This policy forms part of the ToneCoaching Terms of Use. Matters not covered here are governed by those terms.
Effective 18 October 2026 · Version 1.1 · Published by Beijing Jiefang Silin Network Technology Co., Ltd.