Privacy Policy

1. Definition of Personal Information

In this Policy, "personal information" means personal information as defined in Article 2, Paragraph 1 of the Act on the Protection of Personal Information of Japan (the "APPI"). Specifically, it means information about a living individual that can identify a specific individual by name, date of birth or other descriptions, or that contains an individual identification code.

In this Policy, "users" include persons who have completed registration for the Service and persons who have applied for membership of the Service ("applicants").

2. Personal Information We Collect

We collect the following personal information to provide the Service.

2.1 Information Provided Directly by Users

2.2 Information Automatically Collected Through Use of the Service

2.3 Content Created by Users

• Posts (timeline posts, board posts, forum posts)
• Comments and reactions
• Direct messages
• Attached images
• Drafts (posts not yet published)

2.4 Information Related to Use of the Service

* We do not hold payment method information such as credit card numbers; it is managed by our payment processor (Stripe).

2.5 Collection of Public Information

To provide the Service, we may collect information about users from the following public sources.

• Databases published by public authorities (such as the National Tax Agency Corporate Number Publication Site and gBizINFO of the Ministry of Economy, Trade and Industry): company name, corporate number, address, name of the representative, public awards and filings, etc.
• Information publicly available on the web (such as the official website of a company where the user serves as the representative or an officer, press releases and news articles): the company's business, titles, published careers and achievements, etc.

We collect only public information linked to the name and company information registered by the user. We do not collect special care-required personal information (such as race, creed or medical history).

2.6 Information Collected from Applicants

The Service is invitation-only. We collect the following information from persons who wish to join, in order to issue invitations and decide whether to issue them. We may collect it from applicants who have not completed registration for the Service.

If an applicant becomes a member, the applicant's information is carried over as user information. If the applicant does not become a member, the information is deleted after the period set out in Section 7.2.

2.7 Information Collected Through Use of Discover

We collect the following information in the "Discover" feature of the Service.

• History of displays, views and clicks of listed content
• History of operations the user performed, such as obtaining materials, making inquiries and booking schedules
• Identifiers of the source and destination when the user moves from listed content to an external site
• History of the user hiding listed content

2.8 Information Collected When Connecting External Services

When a user connects an external service set out in Section 5.6 (a "connected service") to the Service from "Settings → External services", we collect the following information, limited to the targets the user selected in the Service (by default, everything the user authorized) within the scope the user authorized on the connected service's consent screen.

We do not store the email addresses or account names of connected service accounts.

2.9 Information Collected When Checking "Whether There Are Matching Executives" Before Joining

We provide a feature that lets people considering joining the Service check, before joining, whether there are executives among the users of the Service who match them ("pre-membership check"). The pre-membership check collects the following information.

In the pre-membership check, the input and the website content are processed by the external AI service set out in Section 16.4 (the Anthropic API provided by Anthropic, PBC) and compared with information users of the Service have made public. We display only a short description of users' industries or roles and the reasons they match; we do not display users' names or company names. Results for the same input are stored for 24 hours, linked to a value converted from the input in a way that cannot be reversed, and reused when the same input is received.

2.10 Information Collected Through Use of My Rep

When a user uses My Rep as set out in Chapter 18, we collect the following information.

• The text of the requests the user entrusts to My Rep (what the user is looking for and what the user can offer)
• The user's answers to My Rep's questions
• The user's decisions on proposals made by My Rep, and the user's answers about what happened afterward
• Records of conversations between My Reps, and proposals made from those conversations
• Notes on the user's recent situation recorded by our staff based on the user's public information and the user's communications with us

2.11 Profile Visit History

We store records of visits to other users' profiles (visitor, visited user and time). The visited user can see the number of unique visitors during the last 30 days. Free members see only the count. Rep Premium members and trial users can also see visitors whose settings both at the time of the visit and currently allow their names to be shown.

The default is to show the visitor's name. Users can turn off “Show my name in profile visit history” in settings; their visits then count only toward the total. Visits by users who choose not to record visits are neither recorded nor counted.

2.12 Reading Business Cards and Invitations from Business Cards

1. When a user photographs a business card, we read the image using the external AI service described in Section 16.4 (the Anthropic API provided by Anthropic, PBC) and return the result to the user's device. Neither the image nor the extracted text is stored in our database. We record the time, purpose and processing result category.
2. Invitations from business cards use a shareable link that the user sends from their own device after reviewing the recipient and message. We do not send such invitation emails or SMS messages in our name. To prepare the link, we store the recipient's name, company, title and contact information (email address or phone number) confirmed by the user, the input source, the basis for sending the invitation if entered, the invitation text, and the status and time of link preparation and confirmation. These records are distinct from storing the extracted text as-is.
3. We do not confirm transmission from the device or delivery to the recipient. If the recipient declines invitations, the user must stop sending them. Requests for deletion or other handling of personal information are accepted through the contact described in Section 10.

2.13 Business Information

For each of the user's businesses, we store goals, current circumstances, obstacles, attempts, preferred approaches and what the user can offer, together with sources, the time of the information, whether it is an AI inference or confirmed by the user, and correction history. Information is shared with other users only when the user selects the individual items to share (the default is not to share).

2.14 Membership Badge

A badge beside the name of a Rep Premium member or trial user shows that status to other users. The badge is shown by default, and users can choose to hide it in settings.

3. Purposes of Use of Personal Information

We use the personal information we collect for the following purposes.

However, notwithstanding this Section, we use information obtained from Google (Section 17.5) only for the purposes set out in Section 3.7 (providing user-facing features to the user), and not for Section 3.2 (service improvement and development, statistical processing), Section 3.4 (communications), Section 3.6 or any other purpose.

3.1 Providing and Operating the Service

• Determining free, trial and Rep Premium membership and available features
• Granting, settling and returning AI credits
• Checking eligibility for member benefits and recording applications and use
• Providing profile visit history
• Reading business cards and preparing invitation links for users to share from their own devices

• Creating, managing and authenticating user accounts
• Providing each feature of the Service (posts, messaging, notifications, etc.)
• Providing features for communication between users
• Providing paid services and subscriptions, and processing charges
• Providing the invitation feature and managing invitation quotas
• Accepting membership applications, reviewing eligibility, issuing invitations and contacting applicants about their applications
• Improving the quality of matching and introductions between executives (using the public information we collect as reference information for selecting introduction candidates and generating reasons for introductions)
• User support by our operations team and checks for community operations
• Disclosure to the user (so that the user can check the public information we hold)

3.2 Improving and Developing the Service

• Improving the quality of the Service by analyzing how often features are used
• UI/UX analysis to improve usability
• Planning and developing new features and services
• Statistical processing of usage (carried out after anonymization so that individuals cannot be identified)

3.3 Ensuring Safety

• Detecting and preventing unauthorized access and misuse
• Investigating and responding to violations of the Terms of Service
• Responding to reports and moderation
• Responding to security incidents

3.4 Communications

• Sending guidance on registration procedures and the provision of the Service (including contact by SMS and email)
• Sending important notices about the Service
• Responding to inquiries
• Providing information related to the Service (where the user consents)

3.5 Legal Compliance and Protection of Rights

• Responding as required by law (such as responding to inquiries from investigative authorities)
• Responding to litigation, collecting receivables and otherwise protecting our rights and property

3.6 Providing and Improving Discover

• Delivering listed content and deciding its display order
• Measuring the effectiveness of listings and reporting aggregated results to advertisers
• Reviewing listings and maintaining their quality
• Improving and developing the Discover feature

3.7 Providing the External Service Connection Feature

• To show the user, on the Rep AI screen, consultation suggestions based on information from connected services
• To prepare, in the Rep AI input field, a consultation message with its source when the user taps a suggestion
• To provide, inside Slack, consultations with Rep AI (the Rep app's Home, direct messages and mentions in channels) and the sending of replies the user has confirmed
• To maintain connections, prevent misuse and respond to incidents

3.8 Providing My Rep and the Pre-Membership Check

• To talk with other users' My Reps about the requests the user entrusts to My Rep, and deliver proposals of matching people to the user
• To prepare suggestions of requests, users and listings that may suit the user, and drafts the user can post after reviewing them (including drafts based on public information such as press releases of the user's company)
• To have our operations team look for a counterpart on the user's behalf when a request entrusted to My Rep cannot be fulfilled within the Service
• To show people considering joining whether there are executives among the users of the Service who match them

4. Provision of Personal Information to Third Parties

We do not provide personal information to third parties without the user's consent, except in the following cases.

4.1 Where Required by Law

• Where disclosure is required by law
• Where it is necessary to protect a person's life, body or property and it is difficult to obtain the person's consent
• Where it is especially necessary to improve public health or promote the sound growth of children and it is difficult to obtain the person's consent
• Where it is necessary to cooperate with a national or local government body, or a person entrusted by it, in carrying out duties prescribed by law, and obtaining the person's consent would likely interfere with those duties

4.2 Outsourcing

We may outsource all or part of the handling of personal information to the extent necessary to achieve the purposes of use. In that case, we appropriately supervise the contractor.

4.3 Business Succession

If our business is succeeded due to a merger, company split, business transfer or other reasons, we may provide personal information to the successor in connection with that succession.

4.4 Provision Based on the User's Own Actions in Discover

When a user makes an inquiry, obtains materials or books a schedule through listed content in Discover, we provide the advertiser of that listing with the user's name, title and company name, and the type, date and time of the operation. We make clear in advance, in this Policy, that these operations result in such provision.

For users who have not performed these operations, we do not provide advertisers with any information that can identify individuals, including the fact that they viewed or clicked listed content. Aggregated results of listing effectiveness provided to advertisers are limited to forms that cannot identify individuals (statistics, and breakdowns of segments containing five or more people with the same attributes).

Advertisers may use information obtained under this Section only for responding to the inquiry and other purposes the user can reasonably expect, and must not provide it to third parties as a list or similar.

4.5 Provision Through External AI Services Connected by the User

When a user, at the user's own discretion, connects an external AI service such as ChatGPT or Claude to the Service, information the user can view in the Service (including other users' posts and public profiles) is sent to that external AI service based on the user's instructions. Information sent is handled under the terms and privacy policy of the provider of that external AI service. What we receive is limited to what the external AI service sends when it uses features of the Service (such as the text of a post, criteria for the person the user wants to find, and the text of a request entrusted to My Rep); we never obtain the full conversation between the user and the external AI service.

4.6 Provision to Member Benefit Partners

When a partner checks membership for a benefit, we show only whether a time-limited verification code is valid, the benefit name and the version of its conditions. We do not provide the user's name or other personal information. Information the user enters directly into a partner's booking page or similar service is handled under that partner's terms.

4.7 Partner Applicants

We collect the contact person's name, email address, optional phone number and proposal from businesses applying to offer member benefits, for assessment, communication and agreement on conditions.

5. Integration with External Services

5.1 Cloud Services

* The only personal information provided to Stytch (Twilio Inc.) when connecting external AI services (OAuth) is the email address and the member identifier. Names and other profile information are not provided.

5.2 Payment Services

For purchases through app stores, we receive only purchase verification information from the respective store. Payment method information is managed by the store and is not held by us.

* We do not hold payment information (such as credit card numbers); it is securely managed by Stripe.

5.3 Provision to Third Parties in Foreign Countries [Important]

Some of the external services we use are operated by businesses located in the United States, and users' personal information may be stored on servers in the United States.

The Personal Information Protection System in the United States

The United States does not have comprehensive personal information protection legislation equivalent to the APPI, but it has sector-specific regulations (such as finance and healthcare) and enforcement by the Federal Trade Commission (FTC). For details, please see the website of the Personal Information Protection Commission of Japan.

Protective Measures Taken by Each Business

• Google LLC: certified under the EU-U.S. Data Privacy Framework; SOC 2/3 certified
• Stripe, Inc.: PCI DSS Level 1 compliant; SOC 2 Type II certified
• Cloudflare, Inc.: ISO 27001 certified; SOC 2 Type II certified
• Twilio Inc. (Stytch): ISO/IEC 27001, 27017 and 27018 certified; SOC 2 Type 2 certified
• OpenAI, L.L.C.: we have confirmed, based on public information, that it holds major security certifications such as SOC 2 (for the latest information, see trust.openai.com). Under its data usage policy, data sent via the API is, as a rule, not used for model training.
• Anthropic, PBC: we have confirmed, based on public information, that it holds major security certifications such as SOC 2 (for the latest information, see trust.anthropic.com). Under its data usage policy, data sent via the API is, as a rule, not used for model training.

5.4 Privacy Policies of External Services

For how each external service handles personal information, please see each company's privacy policy.

• Google Privacy Policy
• Cloudflare Privacy Policy
• Stripe Privacy Policy
• Twilio Privacy Policy
• OpenAI Privacy Policy
• Anthropic Privacy Policy

5.5 Search Service Providers

• Search service providers (such as Brave Software, Inc.): we may send search queries that include users' names and company names in order to search public information

5.6 External Services Connected by Users

The following services are connected to the Service by users themselves with their own accounts; they are not our contractors. Information in each service is handled under that company's privacy policy.

The Service does not post, edit, delete or otherwise write to the services above. However, for Slack, as set out in Section 17.7, the Service posts responses and displays as the Rep app, and posts replies the user has reviewed and confirmed.

6. Handling of Direct Messages

1. As a rule, we do not view the content of direct messages (DMs) between users.
2. However, we may check reported messages only in the following cases:
   • When a user reports the message
   • When disclosure is requested under law
   • When necessary to prevent crime or cooperate with an investigation
3. In the cases above, we act in compliance with the Telecommunications Business Act and other related laws.
4. When a user withdraws, DMs sent by that user remain in the recipient's message history, but the sender's name is shown as "Withdrawn user".

7. Storage and Protection of Personal Information

7.1 Security Control Measures

To prevent leakage, loss or damage of personal information, we take the following security control measures.

Organizational Measures

• Appointing a person responsible for personal information protection
• Establishing and operating rules on the handling of personal information
• Recording the handling status and conducting regular inspections and audits
• Establishing a system for responding to incidents such as leaks

Human Measures

• Regular education and training for employees
• Employment contracts and work rules that include confidentiality obligations

Physical Measures

• Physical security for servers and other equipment (locking, access control)
• Measures to prevent theft of devices and electronic media

Technical Measures

• Encryption of communications (HTTPS/TLS 1.2 or later)
• Managing access rights and keeping access records
• Secure storage of credentials (hashing, encrypted storage)
• Systems to detect and prevent unauthorized access (WAF, IDS/IPS)
• Regular vulnerability assessments

7.2 Retention Periods

We retain personal information for the following periods.

8. Handling of Data upon Withdrawal

8.1 Data Deleted Immediately

After the withdrawal procedure is completed, the following data is deleted immediately.

• Credentials (login authentication tokens, session information)
• Drafts (posts not yet published)
• Notification settings and notification history
• Confirmation records for invitations from business cards
• Purpose consent records for business cards and external information (including withdrawal records and pre-registration records of the same person)
• Profile visit records and member benefit application and usage records
• Block and mute lists
• External service connection information (encrypted tokens, selected targets)

8.2 Data Deleted After a Certain Period

The following data is retained for 30 days from the date the withdrawal procedure is completed and then completely deleted.

• Profile information (name, self-introduction, profile image)
• Account information (user ID, registration information)

8.3 Handling of Posted Content

• When a user withdraws, content posted by that user (timeline posts, board posts, comments, etc.) continues to be displayed, with the poster's name changed to "Withdrawn user".
• If a user wishes to delete posted content completely, please delete each post yourself before withdrawing.
• Reactions (emoji) are anonymized and only the aggregated results remain.

8.4 Anonymized Information

For service improvement and statistical purposes, we may create and use "anonymized information", which is personal information processed so that individuals cannot be identified. Anonymized information does not include information that can identify individuals, such as user IDs, names or contact details. We do not use information obtained from Google to create anonymized information.

9. Handling of Special Care-Required Personal Information

1. We do not intentionally collect special care-required personal information (race, creed, social status, medical history, criminal record, the fact of having suffered damage from a crime, etc.).
2. When uploading a profile image, please be careful not to include information from which special care-required personal information may be inferred (such as photos taken at a specific religious facility or at a medical institution).
3. If content containing special care-required personal information is posted, we may delete that content.

10. Users' Rights

Users have the following rights against us.

10.1 Request for Disclosure

Users may request disclosure of their personal information that we hold.

10.2 Request for Correction, Addition or Deletion

If the content of their personal information is not true, users may request its correction, addition or deletion.

10.3 Request to Stop Use or Erase

If their personal information is handled beyond the scope necessary to achieve the purposes of use, or was obtained by wrongful means, users may request that we stop using or erase it.

10.4 Request to Stop Provision to Third Parties

If their personal information is provided to third parties without their consent, users may request that we stop providing it to third parties.

10.5 Procedures for Exercising Rights

How to Make a Request

Please contact us by email at the contact point set out at the end of this Policy. We will send you the prescribed request form.

Identity Verification

Please submit a copy of an official identity document, such as a driver's license, My Number Card (front side only) or passport.

Fees

• Request for disclosure: 1,000 yen (including tax) per request
• Requests for correction, deletion or stopping use: free of charge

Response Time

As a rule, we will respond within two weeks after receiving a request. If an investigation takes time, we will let you know.

Method of Response

By email or by postal mail (as you prefer)

10.6 Checking, Correcting and Deleting Public Information by the User

Users can check, correct and delete, at any time and free of charge, the public information about them that we have collected, from the "Your public information known to Rep" screen in the app. Deleted items will not be restored by subsequent automatic collection.

(The procedures for requests for disclosure, etc. remain available in addition to this screen.)

11. Use of Cookies

11.1 What Cookies Are

Cookies are small text files stored in the browser when a website is viewed.

11.2 Types of Cookies We Use

11.3 Managing Cookies

You can disable or delete cookies in your browser settings. However, if you disable essential cookies, some features of the Service may become unavailable.

How to manage cookies in major browsers:

• Chrome: Settings → Privacy and security → Cookies
• Safari: Settings → Privacy
• Firefox: Settings → Privacy & Security

11.4 Moving to External Sites from Discover

When a user moves from listed content in Discover to an external site, we add an identifier (click ID) and parameters to the destination URL to measure the effectiveness of the listing. This identifier does not directly identify individuals.

12. Use by Persons Under 18

1. The Service is intended for persons aged 18 or older. Persons under 18 may not use the Service.
2. If we find that we have collected personal information from a person under 18, we will suspend the account and promptly delete the personal information collected.
3. If you become aware that a person under 18 may be using the Service, please contact our contact point.

13. Changes to This Policy

1. We may change this Policy due to changes in laws, changes in the content of the Service or other reasons.
2. The changed Policy takes effect when we notify users by posting it in the Service or by other methods we consider appropriate.
3. For important changes, we will notify users in advance, at least 14 days before the change takes effect, by a notice in the Service, push notification, email or other means.

14. Language

This Policy is prepared in Japanese. If there is any conflict or inconsistency between the Japanese version and a version in another language, the Japanese version prevails.

15. Matters Published Under the APPI

15.1 Business Handling Personal Information

15.2 Person Responsible for Personal Information Protection

Contact: same as the contact point below

16. Information Processing by AI Features

In the Service, we may use AI technology to aggregate and summarize information, analyze profiles, and generate an External View for the user.

16.1 Information Covered

16.2 Purposes of Use

• To present to the user an abstracted view of "how other users see you" to support self-understanding.
• To organize the user's posting and interaction tendencies in a way that is easy for the user to understand.
• To improve the quality of AI features, reduce erroneous output and evaluate safety.
• To operate the Service stably, investigate incidents and prevent misuse.

16.3 Safeguards for External View

In generating and displaying External View, we take the following safeguards.

• We do not display individual External View results to users other than the user concerned.
• To avoid displays from which individuals could be inferred from reactions or comments by a small number of other users, we aggregate only when there are observations from at least a certain number of other users.
• We avoid quoting original text, proper nouns and expressions that point to a single person, and abstract the content as tendencies from multiple people.
• We have mechanisms to detect and suppress, before and after display or storage, information in AI-generated results that could lead to identifying individuals.
• We do not use data that has been deleted, hidden, or whose consent has been withdrawn to generate new External Views.

16.4 Use of External AI Services

We also use the Anthropic API provided by Anthropic, PBC to read business cards (Section 2.12).

For AI analysis and generating External View, we may use the OpenAI API, the Anthropic API and other external AI services we select. These services may be provided by businesses located in the United States or other countries outside Japan.

Information sent to external AI services is limited to what is necessary for generation. We check the descriptions of the data usage conditions, retention periods, whether data is used for retraining, subcontracting and cross-border transfers of the external AI services we use, and explain them to users as necessary.

16.5 Consent, Withdrawal and Deletion

1. Generating and viewing External View requires explicit consent to the version of the terms we designate.
2. For users who have consented, we record the version they consented to.
3. If a user withdraws consent, we physically delete the user's aggregated External View results and stop generating and viewing External View until consent is given again.
4. If a user withdraws from the Service, we hide or logically delete AI analysis data at withdrawal and, as a guideline, physically delete it 30 days after withdrawal, except data that must be retained for legal or operational reasons.

16.6 Retention Periods

17. Handling of Information from Connected External Services

17.1 General

1. Connecting a connected service is optional. Other features of the Service remain available without connecting.
2. We use only the targets the user selected in the Service (by default, everything the user authorized) within the scope the user authorized on the connected service. If the user chooses not to select any target, we do not obtain information from that service.
3. We obtain information from connected services each time in response to the user's actions. Titles, content and other information obtained are used to show suggestions and prepare consultation messages, and are not stored in our database. However, when the user uses My Rep, we read connected-service information periodically and store only a one-line summary, as set out in Section 18.5 (the original content is not stored).
4. Access tokens and refresh tokens are encrypted with AES-256-GCM before storage and are used only for that user's own connection.
5. Information obtained from connected services (including today notes under Section 18.5) is shown only to the user and not to other users. Even when today notes created from non-Google connected services are used as material for the user's own My Rep judgments and drafts under Section 18.5(3), they are never shared with other users or their My Reps.

17.2 Use for Rep AI Consultations

1. When the user taps a suggestion, we obtain the latest information from the connected service and prepare, in the Rep AI input field, a consultation message with its source. The message is not sent automatically.
2. Only when the user reviews and edits the consultation message and sends it themselves is it saved as a Rep AI conversation and processed by the external AI service set out in Section 16.4 (the Anthropic API provided by Anthropic, PBC) to generate the reply to the user. We may also convert the consultation message into numerical values with the OpenAI API to look for helpful information within the Service. We do not store the results of this conversion or the text used for the search.
3. Consultations started from suggestions of connected services (including later messages in the same conversation; "externally sourced consultations") are used only for replies to the user and for features the user uses (creating conversation titles, listing and searching conversations, and creating drafts the user requested). We do not use externally sourced consultations for the following purposes:
   • Activity records, updating the user's summary or profile, or extracting business information
   • Building indexes for member search, quoting them as context for other consultations, or as material for reflection features
   • Displaying member details to our staff, analyzing consultations, or including their content in notifications to our operations team
4. We do not use information obtained from connected services or externally sourced consultations for serving or targeting ads, selling to other businesses, or training AI models.
5. How providers of external AI services handle the information sent (including whether it is used for model training) is governed by Sections 5.3 and 16.4.

17.3 Retention and Deletion

17.4 Removing Connections and Revoking Permissions

1. Users can remove a connection at any time from "Settings → External services". When a connection is removed, we delete the stored connection information and selected targets and no longer obtain information from that service.
2. Removing a connection in the Service does not automatically revoke the permissions granted to the Service on the connected service. To revoke them, please use the settings of each service.
3. Even after a connection is removed, externally sourced consultations the user has already sent remain as Rep AI conversations. Requests to delete them are accepted at the contact point in Chapter 19.

17.5 Google User Data

1. Rep's use and transfer to any other app of information received from Google APIs will adhere to the Google API Services User Data Policy, including the Limited Use requirements. The use of information received from Google Workspace scopes will adhere to the Google User Data Policy, including the Limited Use requirements.
2. We access the following data from Google, read-only. We never create, change or respond to events, and never create, change or delete files.
   • Google Calendar: the list of calendars, and, for the calendars the user selected in the Service (by default, all calendars the user authorized), the title, date and time, location, number of attendees, an excerpt of the description and a link of events from now until 7 days later (we neither use nor store the email addresses of attendees or organizers)
   • Google Drive: only the files the user selects in Google's picker (Google Picker) — file name, last modified time, owner's display name, link, and an excerpt of the body of Google Docs (we do not access the owner's email address)
3. To show the Google Picker, we temporarily pass to the user's own browser the access token issued for the user's Google Drive connection, which can only be used for the files the user selects.
4. Notwithstanding Sections 3, 8.4 and 16, we use data received from Google only to provide user-facing features to the user, within the scope of Sections 17.1 to 17.4. Today notes (Section 18.5) created from data received from Google are used only to display them to the user and to generate Rep AI replies to that user; they are not used as material for judging matching counterparts with other users or for drafts. We do not use it for service improvement or developing new services, statistics or analytics, creating anonymized data, sending announcements or marketing, or improving our AI features. We do not use it for serving ads (including retargeting, personalized or interest-based ads), selling it to others, determining creditworthiness or lending. We do not use it to develop, train or improve generalized AI or machine-learning models.
5. We do not transfer data received from Google to third parties except: (1) when necessary to provide user-facing features and with the user's consent (including processing, by external AI services, Google-derived information contained in a consultation the user chose to send, solely to generate the reply to that user, as described in Section 17.2); (2) for security purposes such as investigating abuse; (3) to comply with applicable laws; or (4) as part of a merger, acquisition or sale of assets, with the user's explicit prior consent.
6. Our staff do not read data received from Google except: (1) with the user's explicit consent; (2) for security purposes such as investigating abuse; or (3) to comply with applicable laws.
7. Users can revoke Rep's access at any time from "Third-party apps & services" in their Google Account. Removing the connection in the Service does not remove Google's permission for the files selected in Google Drive, so please use this page to revoke it.

17.6 Microsoft 365 Data

1. We access the following data from Microsoft, read-only. We never create or change events, send, change or delete emails, or change files.
   • Outlook Calendar: the list of calendars, and the title, date and time, location, number of attendees, an excerpt of the description and a link of upcoming events in the calendars the user selected in the Service (we do not access the email addresses of attendees or organizers)
   • Outlook Mail: the list of folders, and the subject, date and time received, sender's display name, an excerpt of the body and a link of unread or flagged emails addressed to the user (we neither display nor store the email addresses of senders or recipients in Rep). We do not use, for suggestions, emails the user sent or automated emails from senders that include unsubscribe instructions or do not accept replies
   • OneDrive: the list of folders, and the name, last modified time, location and link of files the user recently modified, and an excerpt of the body of Word documents (.docx) (the body is read only when the user taps a suggestion and is not stored)
   • To determine whether items are addressed to the user, we read the identifier and email address of the user's own Microsoft account (used only for this determination and not stored).
2. We use data obtained from Microsoft only to provide user-facing features to the user, within the scope of Sections 17.1 to 17.4, and do not use it to train AI models.
3. Microsoft-side permissions can be revoked from the page for managing app permissions in the Microsoft account. For work or school accounts, the organization's administrator can also revoke them.

17.7 Slack Data

1. By connecting their own Slack account to the Service and adding our Slack app "Rep" (the "Rep app") to their Slack workspace, users can use the features set out in this Section. The Rep app is added to the extent permitted by the workspace's administrators.
2. The permissions we use in Slack are as follows:
   • The user's own permissions: searching messages (search:read.public, search:read.private, search:read.im, search:read.mpim) and posting in the user's own name (chat:write)
   • The Rep app's permissions: message shortcuts (commands), posting as the Rep app (chat:write), reading direct messages between the user and the Rep app (im:history), and reading messages that mention the Rep app (app_mentions:read)
3. We search Slack, within the scope of the user's own permissions, only when the user opens, refreshes or searches the Slack topics screen in the Service, and when the user opens the Rep app's Home in Slack. Search results (the sender, channel, an excerpt of the content and a link of messages) are used only to display them on the screen and are not stored in our database.
4. When the user sends a direct message to the Rep app, starts a consultation by choosing a topic on the Rep app's Home, mentions the Rep app in a channel, or presses "メッセージで続ける" (Continue in Messages), we read the following information each time to generate the reply to the user:
   • Messages in the consultation thread between the user and the Rep app (up to the latest 20 messages)
   • The topic selected on Home (searched again with the user's own permissions, and only if it is still found as addressed to the user)
   • The text of the message that mentioned the Rep app in a channel (we do not read other messages in that channel)
5. Rep AI generates these replies using also the user's own information in the Service (profile, business information, etc.), and they are processed by the external AI service set out in Section 16.4 (the Anthropic API provided by Anthropic, PBC). We may also convert the text of the consultation into numerical values with the OpenAI API to look for helpful information within the Service. We do not store the results of this conversion or the text used for the search. Replies to mentions in a channel are shown only to the user who mentioned the Rep app.
6. The content of consultations inside Slack (the user's messages, the selected topics and Rep AI's replies) remains only in Slack; it is not stored in our database and is not added to the history of Rep AI conversations. We store only records showing that a consultation took place (the user, identifiers of the workspace, channel and thread, and the dates and times of the first and last messages) and usage records (counts and processing volume, not including content).
7. In consultations about a topic selected on Home, Rep AI drafts a reply to the original message when the user asks for it. A reply is posted to the original thread in the user's own name only when the user has reviewed the recipient, thread and content and confirmed sending; it is never sent automatically. Right before sending, we confirm that the original message is still found as addressed to the user. To prevent duplicate sending, for each send operation we store a matching value calculated from the recipient, thread and content (a value from which the original content cannot be restored), the processing status, and the identifier of the result. We do not store the content of the draft or of the reply itself. These records are deleted when the withdrawal procedure is completed.
8. Only if the user turns on "Also send Rep notifications to Slack" in the Service settings, we also send the Service's notifications (the same scope and wording as push notifications) to the user as direct messages from the Rep app. The only new information we store for this feature is whether this setting is on.
9. When the user selects "Rep に相談" (Ask Rep) from a Slack message's menu, we keep the channel, message identifier, sender's identifier and display name, and content (the first 4,000 characters) of that message, encrypted and valid for only 5 minutes, and delete it when the user opens it in the Service. If it is not opened, it is deleted within 24 hours after it expires. The content is placed in the Rep AI input field and is saved as a Rep AI conversation under Section 17.2 only if the user sends it themselves.
10. To link users to their actions in Slack, we store a matching value calculated from the identifiers of the Slack workspace and the user. When the Rep app is added, we store the identifier of that workspace, the user who added it and the date and time, as well as the Rep app's credentials (access token and refresh token, encrypted with AES-256-GCM). The credentials are deleted when the Rep app is removed from the workspace or its permissions are revoked. Because the workspace record is also used by other users of the same workspace, it remains after the user withdraws, but the link to the user who added it is deleted when the withdrawal procedure is completed. To discard duplicate notifications from Slack, we store a non-reversible matching value for each notification and delete values older than 1 day the next time we receive a notification.
11. If a Slack user who has not connected to the Service uses the Rep app, we only return guidance on connecting and do not store the content.
12. We do not use information obtained from Slack for serving ads, selling to other businesses, or training AI models.
13. Rep AI's replies may be inaccurate. Please check the original information before making important decisions.

18. Handling of Information by My Rep

My Rep is a feature that, for requests the user entrusts to it, talks with other users' My Reps on the user's behalf and delivers proposals of matching counterparts to the user.

18.1 Information Used for Judgments and Information Shared with Counterparts

Of the business information in Section 2.13, only items the user has selected to share with other users are conveyed to other users' My Reps.

1. To judge whether someone is a matching counterpart, My Rep processes the user's information we hold (profile, business information, key points organized from posts in the Service, notes on the user's recent situation recorded by our staff, press releases published by the user's company, the information in today notes under Section 18.5 (excluding those created from data received from Google), the information under Section 2.10, etc.) with the external AI services set out in Section 16.4.
2. What is shared with other users' My Reps is limited to the profile and business information the user has made public and the text of the request the user entrusted. Information the user has not made public is not shared. Of the information in item 1, notes on the user's recent situation recorded by our staff are not shared with other users' My Reps and are used only to judge whether someone is a matching counterpart.
3. When another user's My Rep approaches the user's My Rep, we likewise use the user's information for judgment and do not share information the user has not made public with the other party.
4. The content of direct messages, and whether, how many times and when messages were sent or received, are not used for My Rep's judgments or conversations.

18.2 Before Reaching the Counterpart

1. Records of conversations between My Reps and the proposals made from them are stored and shown to the user.
2. A proposal does not reach the counterpart user until the user chooses to proceed. My Rep does not make payments or other commitments on the user's behalf.
3. The user can review topics My Rep decided not to pursue and bring them back.

18.3 Requests from External AI Services

Users can entrust requests to My Rep from the external AI services set out in Section 4.5 by the user's explicit instruction. In that case, what we receive is limited to the text of the request entrusted.

18.4 Stopping Use

Users can stop or withdraw requests entrusted to My Rep at any time.

18.5 Today Notes (Summaries of Connected-Service Information)

1. When the user uses My Rep and has connected a connected service set out in Section 5.6, we periodically read information from the connected service (event titles, dates and descriptions; task titles and bodies; the text of messages addressed to the user; etc.), once every 1 to 7 days per user, so that My Rep understands the user's current situation, and summarize it with the external AI service set out in Section 16.4 into one line of "status" and one line of "next step" ("today notes"). We store only the one-line summary, the one-line next step, the type and a link to the source; the original content is not stored.
2. Today notes are shown only to the user. The user can dismiss a today note at any time, and dismissed notes are not shown. Today notes are deleted 7 days after creation, and notes created from a connection are deleted when that connection is removed.
3. Today notes created from non-Google connected services (Slack, Linear, Todoist, Asana, Notion, Chatwork and Microsoft 365) are used as material for the user's own My Rep to judge matching counterparts (Section 18.1(1)), for Rep AI replies to the user, and for drafts of the user's own posts (which are not published until the user reviews and posts them). They are never shared with other users or their My Reps. Today notes containing company names, amounts, personal names or contact details are not used as material for drafts.
4. Today notes created from data received from Google are, as set out in Section 17.5, used only to display them to the user and to generate Rep AI replies to that user; they are not used as material for judging matching counterparts with other users or for drafts.
5. Instructions contained in connected-service content (such as "ignore this") are not followed and are treated as content. Today notes may be inaccurate; please check the original information before making important decisions.

18.6 Limits on My Rep Usage

We manage processing that My Rep performs on its own without the user's action through the AI credit balance (Section 2.4). Automatic processing stops when the balance falls below the daily replenishment amount for the user's membership, and we notify the user. Processing initiated by the user remains available within the remaining balance.

19. Contact

For inquiries about this Policy and requests for disclosure, correction, deletion, etc. of personal information, please contact:

GARO Inc. Personal Information Inquiries

• Email: rep-support@garoinc.jp
• Hours: Weekdays 10:00–18:00 JST (excluding weekends, national holidays and the year-end and New Year holidays)

Revision History