1. Scope & contact
This policy explains how the BAG-BOT service offered through baggoblin.com under the BAG Goblin brand handles information from website visitors, customers, and people interacting with connected Telegram bots, the shared BAG-BOT Discord app, or the business private pilot.
For privacy questions or requests, contact support@baggoblin.com. Community administrators also control their group's or server's settings and content; contact them about how they use the bot in their community.
2. Information we process
- Setup and support requests: your name, email, Telegram or Discord handle, project name, community type and size, plan preferences, selected features, goals, and messages or attachments you choose to send. Website request forms are processed through Netlify Forms.
- Telegram Free enrollment: the setup, owner and enrolled group identifiers, enrollment time and status needed to enforce one group per account, reserve capacity and apply Free access. Free enrollment may remain reserved through pauses and paid upgrades. Paid configuration remains stored for a later upgrade but is not used for Free community-information replies. Free does not send questions to an AI provider.
- Dashboard account: your Telegram or Discord user ID and profile name, sign-in/session records, and information used to verify that you control the bot or administer the group or server.
- Discord sign-in: Discord supplies your basic profile and server list with ownership or permission information so we can show servers you can manage and check your access. We save your linked account, chosen server ID and name, and encrypted OAuth access and refresh tokens. Discord sign-in does not request your email address or password.
- FAQ imports and reviews: submitted questions and answers are validated for a preview and saved only after you apply them. Saved FAQ review records include a content fingerprint and the time you explicitly marked that collection reviewed. Imports are not used to train a model, and the uploaded file is not kept as a separate document.
- Bot configuration: project knowledge, FAQs, official links, community rules, preferences, bot and group/server identifiers, encrypted Telegram bot tokens, and related connection details.
- Telegram activity: message text or captions, sender and group identifiers, join events, commands, and related Telegram metadata delivered to the bot for its configured functions. Submitted feedback includes the text and sender ID.
- Discord commands: interaction, server and member identifiers, command names, submitted questions or FAQ searches, and related metadata delivered by Discord. We keep short-lived records to prevent duplicate replies and enforce usage limits. The Discord service does not read ordinary server chat or message history, and does not monitor joins or provide automatic moderation or welcome messages.
- Billing and access: Telegram Stripe customer, subscription and checkout references, selected plan, trial use, payment status and billing dates; Discord trial dates, Free enrollment and reserved-capacity status, activation and pause status, and usage counts. Stripe processes payment details; the BAG-BOT app does not store full payment-card numbers. Discord has no paid checkout.
- Service reporting: limited setup and delivery milestones associated with a setup or workspace help the service owner see aggregate connections, activations and successful replies. Reports do not add message text, bot tokens or visitor tracking. Existing setup and operational records are still processed for the functions described in this policy.
- Operations: usage counts, anti-spam signals, processing records, security/audit events, and technical information handled by our hosting and email providers to deliver and protect the service.
3. Why we use information
We use information to respond to requests, create and operate your bot, authenticate customers, verify group or server permissions, apply supported Telegram moderation rules, answer requested questions, route Telegram feedback, manage subscriptions, enforce usage and trial limits, troubleshoot problems, prevent abuse, and meet applicable legal obligations.
Project content is used for the configured community. The configured Telegram community owner may see feedback submitted through their bot, including associated sender information. Telegram bot answers and moderation actions may be visible to other group members. Discord command replies use Discord's private, or ephemeral, reply setting: they are shown to the person who ran the command. Discord, BAG-BOT, and the providers described here still process the information needed to deliver those replies.
4. Optional AI processing
When Telegram Pro needs AI generation, the relevant question or draft prompt and selected configured project knowledge are sent to OpenAI. An exact approved FAQ match can be answered directly without an OpenAI request. Telegram Basic does not include AI generation.
Discord AI starts off and requires an unexpired eligible trial. A server administrator can enable it during that trial, after which /bagbot ask can answer an exact approved FAQ directly. When AI generation is needed, the member's question and selected configured server information are sent to OpenAI, along with a hashed identifier derived from the workspace and member for safety purposes. Saved FAQ, rules, links, about, help, and support commands do not call OpenAI.
The dashboard service does not maintain a conversation-history feature for AI prompts. It requests that generated responses not be stored as retrievable response objects. Questions and generated answers can be held briefly in the processing queue to complete a reply. This does not mean OpenAI has zero retention: its security, abuse-monitoring, and other applicable retention practices still apply. See OpenAI's API data controls.
Earlier or separately hosted Telegram setups may also send recent direct conversation context to the AI provider and may use the provider's default response-storage behavior. Contact support to confirm which setup applies to your community.
Do not put passwords, bot tokens, payment details, health records, private keys, or other sensitive information in AI prompts or project knowledge. Administrators should use public community information and tell members when AI is enabled. Members can use the saved-information commands without using AI.
5. Providers & other disclosures
Information is shared with providers as needed for their role:
- Telegram: sign-in, bot management, and delivery of group interactions.
- Discord: sign-in, server permission checks, app installation, and delivery of commands and private replies. Discord also processes information under its own Privacy Policy.
- Stripe: checkout, payments, billing, and subscription management. Account and setup identifiers may be included to match a payment to the correct service.
- OpenAI: optional AI responses and drafts as described above.
- Render: hosting and data storage for the dashboard service.
- Netlify: hosting the public website and processing setup forms.
- Retell: operation of the separate private telephone pilot, including call audio, transcripts, extracted callback requests and caller metadata as described below.
- Resend: delivery of generic callback notices to the fixed owner email address when alerts are enabled. Those notices do not include caller names, phone numbers, transcripts or recordings.
- Email providers: delivery and storage of support correspondence.
These providers may process information in countries outside your own under their applicable terms and privacy practices. We may also disclose information where legally required or reasonably necessary to investigate abuse, secure the service, or protect rights.
This website and dashboard do not use advertising trackers or sell personal information through the service, and do not share it for cross-context behavioral advertising.
6. Cookies & website demo
The website demo runs in your browser; demo input is not sent to BAG-BOT's AI service. Submitting the setup form does send the form data to Netlify. BAG-BOT does not add advertising cookies to the marketing site. Netlify hosting can add performance-monitoring code that measures page loading, responsiveness, and layout stability and, when allowed by browser security settings, sends those measurements to Netlify. Netlify describes this measurement as cookie-free. Its hosting badge can also use local browser storage to remember whether you dismissed it; that preference is not advertising tracking. See Netlify's monitoring documentation and badge documentation.
The dashboard uses an essential cookie to keep you signed in. Blocking it may prevent login. Third-party sites such as Telegram, Discord, and Stripe use their own cookies and privacy practices when you visit them. Hosting providers may process technical request data.
This website has no cross-site advertising tracking to change in response to a browser Do Not Track signal and does not use that signal to alter essential service functions.
7. Retention & security
We keep account details, bot settings, billing references, and support records as needed to operate the service, handle requests or disputes, prevent abuse, and satisfy applicable recordkeeping obligations. Those records do not all have automatic deletion timers. Discord workspace settings, Free enrollment and sign-in grants remain until removed through support; trial expiry, pausing or signing out does not delete them.
Limited first-setup and reply milestones and saved FAQ review metadata remain associated with the setup while needed for the service and are included in the verified deletion process below. Aggregate reporting describes observations since the reporting feature was introduced; it does not reconstruct earlier message history.
In the dashboard service, message payloads used by the processing queue are normally cleared when processing finishes or permanently fails. Discord questions, member identifiers, reply tokens, and generated answers are encrypted while in that queue. Discord reply jobs have a 14-minute delivery deadline, after which routine cleanup clears expired pending payloads. Duplicate-response records, which may contain saved-information replies, expire after 24 hours. They do not contain a retained AI conversation history.
Routine cleanup removes expired sign-in flows and sessions, older processing records, Telegram feedback older than 90 days, and audit records older than 180 days. Cleanup depends on successful system operation, so these are normal operating targets rather than a guarantee of deletion at an exact time.
Usage records and limited account/group or server identifiers used to prevent repeated trials may remain after a trial ends. Discord API data is subject to the deletion process below; trial-abuse prevention is not a blanket exception to a valid Discord data-deletion request. Provider logs, email, payment records, and backups can follow different retention schedules. Account closure does not immediately erase all copies or records.
Telegram bot tokens, Discord OAuth credentials, and short-lived Discord reply credentials are encrypted in the dashboard database. Access to setup records is restricted through account and permission checks. Earlier separately hosted Telegram setups keep connection credentials in hosting environment settings. No online service can guarantee absolute security. If you suspect a leaked credential, revoke it through the relevant provider and contact support.
8. Your choices & requests
You can change supported project settings in the dashboard, switch Discord AI off, and pause bot responses. Administrators can remove the bot from their group or server. You can revoke BAG-BOT's Discord sign-in access in your Discord account's Authorized Apps settings. These actions do not automatically delete BAG-BOT's existing records or cancel a separate paid Telegram subscription.
Email support@baggoblin.com to request access to, correction of, or deletion of information relating to you. For Discord requests, include your Discord user ID and, if relevant, the server ID and approximate command time so we can locate the records. Never send a password, bot token, or sign-in token. You do not need to own a server to request deletion of your own information.
We may verify your identity or authority before acting. We will promptly address Discord data correction and deletion requests, including deletion of relevant account, workspace, authorization, and operational records as appropriate to the verified request, except where retention is required by law. We also remove Discord API data when it is no longer needed for the service or the Discord service is discontinued. Other records may need to be retained for legal, security, fraud-prevention, or billing reasons where permitted.
Account deletion currently requires support assistance; a dashboard configuration export is not a complete personal-data export. Deletion from BAG-BOT does not delete your Discord or Telegram account, remove data held independently by those services, or cancel Telegram billing. We will explain any records that must be retained and any provider or backup limitations that apply.
Depending on where you live, you may have additional rights, including objection, restriction, portability, or the right to complain to a privacy regulator. We will handle requests as applicable law requires.
9. Children & policy updates
BAG-BOT is offered to people authorized to manage communities and enter a service agreement. The service is not directed at children under 13 or under the higher minimum age required in their country. Discord users must also meet Discord's age requirements. If you believe a child has provided personal information improperly, contact us so we can investigate and address it.
We may update this policy as the service changes. The revised date will appear at the top. Material changes will be communicated where required by applicable law.
10. Business private pilot
The business pilot is access-by-approval only. It reads connected Gmail or Outlook Inbox messages and supports saved email or CSV imports for a follow-up tracker. This mailbox workflow does not send replies, move, delete, or mark mailbox messages as read; attachments are ignored. The separate telephone-pilot notices described below do not use connected inboxes to send mail. Current business suggestions use rules and reply templates, without sending business email to an AI model.
Google user data: BAG-BOT Business follows the Google API Services User Data Policy, including its Limited Use requirements, when using or transferring information obtained through Google APIs. The business email workflow does not send this Google user data to an AI model or use it to train AI models; its suggestions come from rules and reply templates.
Information processed: sender details, received time, subject, message body, category, priority, status, follow-up date, notes, reply drafts, and review notes. We also store the linked account’s identifiers and email, granted permissions, encrypted OAuth access and refresh tokens, mailbox checkpoints, message identifiers used to avoid duplicate imports, destination details, and operational status. Fictional sample previews stay separate from saved customer records.
Connected providers and copies: Google or Microsoft supplies authorized mailbox or file access. Choosing Google Sheets or Excel creates one active dedicated tracker. Sync copies all saved tracker records, including email bodies and drafts, into that file and overwrites its tracker cells; spreadsheet edits do not flow back. Long cell contents are shortened with a notice, so the spreadsheet is not a complete copy of every original message. Google Sheets uses access to files created or opened through the app. Microsoft’s delegated file permission is broader than a single workbook, although this workflow uses its dedicated tracker. Review the provider’s consent screen before granting access.
Storage and encryption limits: saved tracker content, OAuth tokens, mailbox checkpoints, and destination details are encrypted in BAG-BOT’s database on its hosting service. Some account identifiers and operational metadata are not encrypted as those fields. This is not end-to-end encryption: BAG-BOT decrypts protected fields to provide the workflow. CSV and JSON exports, provider spreadsheet copies, and downloaded database backup files are not encrypted by this application as whole files. Protect those copies and their sharing settings separately. A tracker JSON backup contains records; it does not restore account connections or the full service.
Pausing, disconnecting, and deletion: automatic updates remain paused in the private pilot. Pausing, signing out, or disconnecting does not delete saved tracker records, downloads, or cloud spreadsheets. Disconnecting removes the stored account grant and stops future use after in-flight requests settle; provider-side access can also be revoked in that provider’s account settings. Saved tracker records have no automatic age-based deletion in the current pilot. For an access, correction, or deletion request, use business support; we may need to verify identity or authority. Provider accounts, downloads, and backups are separate copies with their own controls, so removing records in one place does not erase every copy.
Separate private telephone pilot: Retell processes the caller’s voice and call content to run the conversation and extract a requested callback. The opening message discloses recording. The current private agent is configured to retain recordings, transcripts, extracted callback details and caller metadata in Retell for seven days. This configured period does not describe every provider security log, support record or separately retained copy. The pilot asks for invented conversational details during testing; the actual voice and originating caller number can still be processed. A requested callback must pass the saved permission and phone-readback checks before it is eligible for an owner notice.
Callback notices and receipts: when enabled, Retell sends BAG-BOT only the validated call and agent identifiers, a schema version, callback status and three eligibility flags in a signed request. The alert integration does not send the caller’s name, number, summary, transcript or recording to BAG-BOT or Resend. BAG-BOT keeps hashed event receipts and delivery-status metadata to avoid duplicate notices; these receipts do not currently have automatic expiry. Resend receives the configured sender and owner recipient addresses and a generic message linking the owner to Retell. A provider’s acceptance of a notice is not proof of inbox delivery or that a callback has been completed. Support requests, provider records and retained backups have separate handling and require review for a data request.