Privacy Policy

This is a translation. The Japanese version is the governing text; in case of any discrepancy, the Japanese version prevails.

Pochiren (the "Operator") sets out below how users' information is handled in the app "Pochiren — Standard Japanese Accent Practice" (the "App") and on the official website pochiren.com (the "Website").

Effective date: October 23, 2026
Last revised: October 26, 2026

Basic principles

• The App does not send your recorded voice or your practice records to the Operator (except when you choose to send a report email from the test version).
• The Website's pronunciation trial page does not send your recorded voice or its results to the Operator either.
• The Website contains no advertising and no analytics that track how you use it.
• The only information the Operator receives is the Website's access logs and the inquiry and report emails you send us.

Part 1: The App

The operator of Pochiren — Standard Japanese Accent Practice (the "App") (trade name: Pochiren; the "Operator") sets out below how users' information is handled in the App.

1. Basic Approach

The Operator does not collect information that can identify users. However, inquiries and reports that users send on their own (Chapter 9) are excluded. Practice data in the App is kept only on the user's device, and the Operator does not receive it.

2. Permissions the App Uses

The App uses only the following two Android permissions.

• Microphone (RECORD_AUDIO): to record pronunciation and score it
• Vibration (VIBRATE): to notify users of actions and results by vibration

The App does not use the Internet permission. The App itself has no feature that communicates over a network. However, the purchase of a paid plan is processed through the Google Play app on the user's device.

3. Recorded Audio

• Recorded audio is used only for scoring, in the device's memory.
• Audio is not saved.
• Audio is never sent outside the device.
• Speech recognition uses only the recognition feature built into the device. On devices that cannot recognize speech on the device, this feature is not shown.
• As an exception, in the test version for users participating in testing on Google Play, only when the user presses the report button and agrees to the content, a recording of the user's comments (the user's voice) and a screenshot of the screen are temporarily saved on the device (they are deleted automatically within 24 hours) and handed to the user's email app. The user decides whether to send them, and the sending itself is performed by the email app. The App itself does not communicate over a network. The production version does not have this feature.
• Also only in the test version, there is a "usage log." The following is recorded on the device: the screens opened, the buttons pressed, whether recording and scoring succeeded (the number of correct, incorrect, and on-hold phrases in the result), the sentences practiced (the sentence number and text), the result of the microphone permission request, the reason a recording stopped or scoring was put on hold, the type of any error and where in the App's code it occurred, and the time. The content of recordings, text the user typed, and information that identifies the user are not recorded. The log is stored only on the device, and entries older than 7 days are deleted automatically. It is also removed when the app is uninstalled. You can delete all of it at any time with the "Delete log" button on the screen shown when you choose the report type "Usage log" with the report button. The App itself does not communicate over a network; the log file reaches the Operator through the user's email app only when the user chooses the report type "Usage log" with the report button and sends it (no screenshot or comment recording is attached). The production version does not have this feature.

4. Information Stored on the Device

The following information is stored only on the user's device. The Operator does not receive it.

• Practice history: the words and sentences practiced, their readings, scores, judgment results, the pitch value of each mora (in semitones), and date and time
• Intonation history: the sentences practiced, the judgment result for the sentence ending, and date and time
• Voice pitch range: a numeric value in frequency (Hz). It is saved for each recording and deleted automatically after 30 days
• App settings: dominant hand, vibration, language, animation, volume, power saving
• Leaves and backgrounds: the number of leaves, the backgrounds you have exchanged for, and the background you have selected

5. How to Delete Information

• You can delete your practice history with "Delete all practice history" on the settings screen.
• Clearing the app's data or uninstalling the app deletes all information on the device.

6. About Android Backup

The App has Android's backup feature (allowBackup) enabled. As a result, the on-device information described in Chapter 4 may be included in the backup of the user's Google account or in data transfers when changing devices. How it is handled depends on Google's policies. The Operator does not receive any of it.

7. Advertising and Analytics Tools

The App contains no mechanism for advertising, usage analytics, or crash-report collection. However, the test version keeps the on-device usage log described in Chapter 3. The log is never sent off the device automatically; we receive it only when you choose "Usage log" in a report and send it.

8. Provision to Third Parties

Except where required by law and for processing by the service providers described in Chapter 9, user information is not provided to third parties (in any case, the Operator does not receive the information in Chapter 4).

9. Inquiries and Reports

• Please do not write health, disability, or other information requiring special care, or other people's personal information, unless it is necessary. If you are under 18, please get your parent's or guardian's consent before sending.
• If you send an inquiry or report by email, your email address and its contents (including attachments) are used only to confirm the content of the report, reply to the inquiry, investigate the cause of problems, and fix and improve the App. Attached recordings, screenshots, and "usage log" files are used only to the extent necessary to confirm the report and investigate the cause of problems.
• The Operator uses the following services as service providers.
  • Receiving and storing email: Google Workspace (provided by Google Asia Pacific Pte. Ltd., Singapore). The country of the contracting party (Singapore) is separate from the countries where data is stored and processed. Storage and processing are not limited to a specific country and take place in countries where Google or its sub-processors have facilities, so the Operator cannot identify the country of storage individually. The countries and regions where Google has data centers can be checked in Google's public information (https://www.google.com/about/datacenters/locations/). This list shows the locations of Google's own data centers; it is not a list of every country where processing takes place, including by sub-processors.
  • Sorting incoming email and drafting replies: Anthropic's Claude API (provided by Anthropic, PBC, United States). Email is passed only if you agree on the App's screen before contacting us. When you agree, [AI consent] (in the Japanese-language App, [AI同意]) is added to the email's subject line, and only emails with this mark are passed. If you do not agree, and for emails without the mark (including those sent by entering the website or address directly), the email is not passed to the AI and the Operator reads it and responds. If the body contains words suggesting information that needs special care, such as illness or disability, the email is not passed to the AI even with your agreement, and the Operator reads it. Only the subject and body of the email are passed; attached recordings, screenshots, and "usage log" files are not passed. Before passing, information detected as identifying, such as email addresses and phone numbers, is deleted or replaced. However, information that could identify an individual from the content or context of the body may remain. The Operator uses settings that specify the United States as the region for API processing (inference) and also limit the storage region to the United States. The scope of and exceptions to these settings are checked in accordance with Anthropic's public materials and contract terms. Anthropic states in its contract terms that it does not use content passed through the API to train AI. Content sent to the API is, by default, not stored after processing. However, content automatically flagged as a safety issue may be stored for up to 2 years.
  • Personal information protection in the United States: there is no comprehensive national personal information protection law; protection comes from laws for specific fields and from state laws (such as the California Consumer Privacy Act). For details, see the survey by the Personal Information Protection Commission of Japan (https://www.ppc.go.jp/personalinfo/legal/kaiseihogohou/#gaikoku).
  • The server that runs the sorting mechanism: a service of Xserver Inc. (Japan).
  • Storing development records: a private repository on GitHub (provided by GitHub, Inc., United States).
• Of the service providers located outside Japan, for Google the Operator uses it after checking the terms of its Cloud Data Processing Addendum, security control measures, sub-processing conditions, and deletion conditions, and confirming that it is ensured that the provider handles information only within the scope of the Operator's instructions. Information is passed to Anthropic only with your consent, and the Operator checks the terms of Anthropic's Data Processing Addendum.
• The Operator checks the implementation status of each company's measures and the foreign systems that may affect them once a year, and also when it learns of an amendment to an agreement, a change of sub-processors, or an event that may affect protection. If a problem arises with the measures, the Operator takes the necessary action, and if they cannot be continued, the Operator stops providing information. You can check each country's personal information protection system in "Survey on Personal Information Protection Systems in Foreign Countries," published by the Personal Information Protection Commission of Japan (https://www.ppc.go.jp/personalinfo/legal/kaiseihogohou/#gaikoku).
• Official materials on each company's contract terms are as follows.
  • Google Cloud Data Processing Addendum: https://cloud.google.com/terms/data-processing-addendum
  • Anthropic Data Processing Addendum: https://www.anthropic.com/legal/data-processing-addendum
  • Anthropic API data retention: https://platform.claude.com/docs/en/manage-claude/api-and-data-retention
  • Anthropic data residency: https://platform.claude.com/docs/en/manage-claude/data-residency
• If you would like to know more about these measures, please contact the address in Chapter 12.
• For matters that need improvement, the Operator may record the necessary text content in its development records (GitHub) after removing identifying information such as names, email addresses, and phone numbers, and confirming that no individual can be identified even from the content or context of the text. Recordings and images are not included.
• The Operator records the date on which the matter is resolved and, from that date, deletes email after one year and files attached to test-version reports (recordings, screenshots, and usage log files) after 90 days. Even after the Operator deletes data, it may take up to 180 days, as set by the service providers, for the data to be completely removed from their systems, after a period during which it can be restored from the trash or by an administrator. Deletion by the Operator and complete deletion by the service providers are separate.

10. Age Requirement

The App is intended for users aged 13 and over. Users aged 13 to 17 must use the App with the consent of a parent or guardian. Users under 13 may not use the App.

11. Changes to This Policy

When this policy is changed, we will notify you in the App or by another method set by the Operator. If you continue to use the App after a change, you are deemed to have agreed to the change.

12. Operator and Contact

• Operator: Pochiren
• Contact: support@pochiren.com

Part 2: The Website

1. Access logs

When you open the Website, the following information is recorded on the server.

• Your IP address and the country estimated from it
• The date and time, and the URL of the page you opened
• Your browser and device type (user agent)
• The page you came from (referrer, if your browser sends it)

This information is used only to investigate faults, respond to unauthorized access and attacks, and keep the Website running reliably. The logs are deleted automatically after 90 days. However, if they are needed to investigate unauthorized access, we may keep them until the investigation is finished.

2. Cloudflare

To deliver the Website quickly and securely, we serve it through the services of Cloudflare, Inc. (United States). Because all access to the Website passes through Cloudflare, Cloudflare also handles information similar to the "Access logs" above. Cloudflare may use necessary cookies for security purposes, such as preventing attacks. For details, please seeCloudflare's Privacy Policy.

3. Cookies, analytics, and advertising

The Website itself does not use cookies. It contains no analytics tools and no advertising.

4. Microphone on the pronunciation trial page

• On the Website's pronunciation trial page, your browser asks for permission to use the microphone so that your voice can be recorded as you read aloud. Even if you do not allow it, you can use the other pages as usual.
• Your recorded voice is processed only within the browser on your device, and is used only to show the tendencies of your pitch accent.
• Neither your voice nor the results shown are saved, and they are never sent outside your device, including to the Operator. They are discarded when you close the page.
• The program that processes your voice is delivered from this Website. No other company's service is used.

Part 3: Inquiries and Reports

Inquiries and reports, whether they come from the App or the Website, are handled as described in Chapter 9 of Part 1. The same text as Chapter 9 is repeated below.

• Please do not write health, disability, or other information requiring special care, or other people's personal information, unless it is necessary. If you are under 18, please get your parent's or guardian's consent before sending.
• If you send an inquiry or report by email, your email address and its contents (including attachments) are used only to confirm the content of the report, reply to the inquiry, investigate the cause of problems, and fix and improve the App. Attached recordings, screenshots, and "usage log" files are used only to the extent necessary to confirm the report and investigate the cause of problems.
• The Operator uses the following services as service providers.
  • Receiving and storing email: Google Workspace (provided by Google Asia Pacific Pte. Ltd., Singapore). The country of the contracting party (Singapore) is separate from the countries where data is stored and processed. Storage and processing are not limited to a specific country and take place in countries where Google or its sub-processors have facilities, so the Operator cannot identify the country of storage individually. The countries and regions where Google has data centers can be checked in Google's public information (https://www.google.com/about/datacenters/locations/). This list shows the locations of Google's own data centers; it is not a list of every country where processing takes place, including by sub-processors.
  • Sorting incoming email and drafting replies: Anthropic's Claude API (provided by Anthropic, PBC, United States). Email is passed only if you agree on the App's screen before contacting us. When you agree, [AI consent] (in the Japanese-language App, [AI同意]) is added to the email's subject line, and only emails with this mark are passed. If you do not agree, and for emails without the mark (including those sent by entering the website or address directly), the email is not passed to the AI and the Operator reads it and responds. If the body contains words suggesting information that needs special care, such as illness or disability, the email is not passed to the AI even with your agreement, and the Operator reads it. Only the subject and body of the email are passed; attached recordings, screenshots, and "usage log" files are not passed. Before passing, information detected as identifying, such as email addresses and phone numbers, is deleted or replaced. However, information that could identify an individual from the content or context of the body may remain. The Operator uses settings that specify the United States as the region for API processing (inference) and also limit the storage region to the United States. The scope of and exceptions to these settings are checked in accordance with Anthropic's public materials and contract terms. Anthropic states in its contract terms that it does not use content passed through the API to train AI. Content sent to the API is, by default, not stored after processing. However, content automatically flagged as a safety issue may be stored for up to 2 years.
  • Personal information protection in the United States: there is no comprehensive national personal information protection law; protection comes from laws for specific fields and from state laws (such as the California Consumer Privacy Act). For details, see the survey by the Personal Information Protection Commission of Japan (https://www.ppc.go.jp/personalinfo/legal/kaiseihogohou/#gaikoku).
  • The server that runs the sorting mechanism: a service of Xserver Inc. (Japan).
  • Storing development records: a private repository on GitHub (provided by GitHub, Inc., United States).
• Of the service providers located outside Japan, for Google the Operator uses it after checking the terms of its Cloud Data Processing Addendum, security control measures, sub-processing conditions, and deletion conditions, and confirming that it is ensured that the provider handles information only within the scope of the Operator's instructions. Information is passed to Anthropic only with your consent, and the Operator checks the terms of Anthropic's Data Processing Addendum.
• The Operator checks the implementation status of each company's measures and the foreign systems that may affect them once a year, and also when it learns of an amendment to an agreement, a change of sub-processors, or an event that may affect protection. If a problem arises with the measures, the Operator takes the necessary action, and if they cannot be continued, the Operator stops providing information. You can check each country's personal information protection system in "Survey on Personal Information Protection Systems in Foreign Countries," published by the Personal Information Protection Commission of Japan (https://www.ppc.go.jp/personalinfo/legal/kaiseihogohou/#gaikoku).
• Official materials on each company's contract terms are as follows.
  • Google Cloud Data Processing Addendum: https://cloud.google.com/terms/data-processing-addendum
  • Anthropic Data Processing Addendum: https://www.anthropic.com/legal/data-processing-addendum
  • Anthropic API data retention: https://platform.claude.com/docs/en/manage-claude/api-and-data-retention
  • Anthropic data residency: https://platform.claude.com/docs/en/manage-claude/data-residency
• If you would like to know more about these measures, please contact the address in Chapter 12.
• For matters that need improvement, the Operator may record the necessary text content in its development records (GitHub) after removing identifying information such as names, email addresses, and phone numbers, and confirming that no individual can be identified even from the content or context of the text. Recordings and images are not included.
• The Operator records the date on which the matter is resolved and, from that date, deletes email after one year and files attached to test-version reports (recordings, screenshots, and usage log files) after 90 days. Even after the Operator deletes data, it may take up to 180 days, as set by the service providers, for the data to be completely removed from their systems, after a period during which it can be restored from the trash or by an administrator. Deletion by the Operator and complete deletion by the service providers are separate.
• A report email is an email sent from the report button in the test version. It contains information the App adds automatically (the App's version, the device model name, the Android version, the screen where the report was made, the sentence being practiced, and the date and time), as well as any recording of your comments and screenshot you attach. A report of the type "Usage log" has the usage log file (Chapter 3 of Part 1) attached instead of a recording or screenshot.
• If you would like your email deleted, please let us know at our contact address.

About businesses located outside Japan

Cloudflare, Inc., Anthropic, PBC, and GitHub, Inc. mentioned above are businesses in the United States, and Google Asia Pacific Pte. Ltd. is a business in Singapore (where email is stored is as described in Chapter 9 of Part 1). The United States does not have a personal information protection system equivalent to Japan's. The Operator uses these companies after confirming that each has measures in place to protect personal information (such as certification under international standards and contractual terms on data handling). For information on the system in the United States, please seethe Personal Information Protection Commission's page on foreign legal systems (in Japanese).

How we protect information

• All communication with the Website is encrypted.
• Server logins use keys only, and the delivery and email admin consoles are protected with two-step verification.
• Only the Operator can view the access logs.
• We collect only the information we need.

Provision to third parties

Except where required by law, we do not provide users' information to third parties. We entrust Cloudflare, Google, Anthropic, and GitHub above with delivering the Website, storing emails, sorting inquiries and reports and drafting replies, and recording improvement points; this is "entrustment" and not provision to third parties.

Requests for disclosure, correction, deletion, and similar

If you wish to request disclosure, correction, deletion, or suspension of use of your information held by the Operator, please contact us at our contact address. After confirming your identity, we will respond promptly.

Changes to this policy

When we change this policy, we will announce it on this page. For important changes, we will also announce it in the App.

Operator and contact

Operator

Pochiren

Operator's name and address

Disclosed without delay upon request to our contact address.

Contact

support@pochiren.com