Privacy Policy
Effective
AI Ticket Bot (“we”, “us”, “our”) is a Discord ticket bot and dashboard. This policy explains what data we collect, what we do with it, and what we never do with it.
We do not sell your data. We do not train external models on it.
1. Who is responsible for your data
AI Ticket Bot is currently operated as an independent project by a sole operator based in Romania. The operator acts as the data controller for personal data processed through the bot, the dashboard, and the marketing site. The operator is contactable at the email addresses below; postal address will be published when a registered legal entity is incorporated, and this section will be updated accordingly.
For privacy questions, data-rights requests, or any concern about how your data is handled, contact us at [email protected] (preferred for support and data-rights) or [email protected] (legal notices, business matters).
2. Data we collect
When you sign in via Discord
We authenticate you with Discord OAuth. Discord returns to us:
- Your public Discord user ID, username, display name, and avatar hash.
- A short-lived access token used solely to fetch the list of Discord servers where you have admin permissions, so we can show them in the dashboard.
We do not receive or store your Discord email, phone number, or password. We do not receive a list of every server you are a member of, only those where the Discord API tells us you have administrative scope.
Server configuration
When a server admin uses the bot, we store the configuration they create: panels, categories, tags, permission rules, embeds, blacklist entries, and audit-log rows that record administrative actions for accountability.
We also record the Discord user ID of the person who added the bot to a server, and of the person whose action started that server’s free trial. We use these to know who to send the service messages described below, and to answer “who set this up?” when a server’s admins ask us.
Messages the bot sends you
The bot sends a small, fixed number of one-time direct messages to the admin identified above. There is no recurring newsletter and no marketing list:
- An introduction when the bot joins your server, explaining how to set it up. If your direct messages are closed, it posts this once in a channel it can write to instead.
- One reminder, roughly three days later, if the server still has no ticket panel. This one is sent by direct message only, never posted in your server, and is never repeated.
- Two messages about the free Premium trial: one shortly before it ends and one when it does, naming the features that switched off.
Each of these is sent at most once per server. Removing the bot stops all of them. We keep a timestamp recording that each was sent, which is what guarantees it is not sent twice; that record is kept even when the message could not be delivered, precisely so we do not retry.
Separately, the person who opened a ticket receives a direct message with the transcript link when it closes, and the recipient of a paid subscription receives a confirmation when it starts.
Referrals
If a buyer names another Discord server as a referral at checkout, we store that server’s ID against the purchase so the reward can be issued and so a reward is not issued twice for the same payment. We do not tell the referred server who named it, and we do not share the buyer’s identity or payment details with it.
Ticket content and brain data
When a ticket is opened, we store the conversation transcript so server staff can review it, the AI can refer to it, and a human can audit it later. Transcripts shared through this site are reachable only via tokenized links and only after signing in as the person who opened the ticket or as the server’s staff; they are never publicly listed.
When your staff teaches the bot something with /ai train (the “brain”), the input is sent once to Anthropic’s API (or to your server’s own AI provider, if it has connected one, see section 3), which rewrites it into a normalized, structured form that the AI can use efficiently on later tickets. Only the rewritten, structured form is stored on our infrastructure; the original wording your staff typed is not retained. This is how brain knowledge persists across tickets without us holding raw staff training text.
Brain content and ticket transcripts sit on our own infrastructure. Anthropic sees ticket content at inference time and training input at the moment of the rewrite, both on a per-request basis with no retention on Anthropic’s side. Two other services see a narrow part of it, described in section 3: Voyage AI receives the brain’s entries to keep them searchable, and Brave receives web search terms. A server that connects its own AI provider sends its tickets to that provider instead of to Anthropic.
Website chat widget visitors
On Premium plans and up, a server owner can embed our chat widget on their own website. When a visitor sends a message, the AI answers it, and if a human is needed the chat hands off to the server’s staff, in a private Discord thread or the dashboard. A web chat is handled as a real support ticket, so the same rules as any other ticket apply (see the ticket and AI sections above): the conversation is stored while it is open, a transcript may be saved when it closes, and it follows the server’s retention and deletion settings.
This site stores nothing about a widget conversation. Your browser keeps only a pointer to the chat (a reference id and a random, non-identifying session id) and a local copy of the visible history, on the website you are visiting. The conversation itself, and any transcript, live with the ticket, keyed to that session id. You can wipe the local copy at any time from the widget’s settings menu, and ask us to erase the stored conversation the same way.
The widget is anonymous by default and sets no advertising or tracking cookies. A server owner may optionally turn on “Sign in with Discord”, which lets a visitor sign in to keep their own chat history and lets the server’s staff see who they are; it is never required to use the widget. A visitor may also leave an optional email or handle so the team can follow up. We send no email from the widget, and use the contact only to reach back about that conversation.
Billing data
If you subscribe to a paid plan or buy a token pack on this site, the payment is handled by Stripe for card and bank payments (under Stripe Managed Payments, Stripe is the merchant of record: it collects any tax due and issues your receipt) or by NOWPayments for cryptocurrency. We never see your full card number or wallet private key. We do receive and store the customer ID, plan code, subscription state, the card brand and its last four digits, invoice IDs, the amount and currency actually charged, and the billing country Stripe determined for tax purposes.
If you subscribe inside Discord instead, Discord is the seller and handles the payment end to end. We never see your payment method, your billing address, or what you were charged: Discord tells us only that your server is entitled to a plan. What we store for those is the Discord subscription ID, the plan, the period, and the price we ourselves set on the Discord listing, which is what our records show as the amount. Cancelling and refunding happen in Discord, under your own account settings.
Usage and analytics
On the public marketing site (aiticketbot.com) we use Google Analytics 4, and only after you grant Analytics consent through the cookie banner. We do not run analytics on the dashboard or inside the bot.
3. Where AI fits in (NEXUS, Anthropic, and your own AI)
Our AI commitment
Unless your server has connected its own AI provider (see “If your server connects its own AI” below), Anthropic’s API touches your content in two situations, both per-request and with no retention on Anthropic’s side:
- At inference time. When a ticket is open, the conversation and the relevant brain entries are sent to Anthropic so the model can generate a reply.
- At training time. When your staff runs /ai train, their raw input is sent once to Anthropic to be rewritten into the normalized brain entry we store. We do not retain the staff’s original wording.
We use Anthropic’s standard API with the default no-training posture: Anthropic does not retain content from API traffic and does not use it to train their models. Anthropic is the only model provider we choose to route inference through. We do not send your data to OpenAI, Google, Microsoft, Meta, or any other AI provider of our own accord; the one way a different model provider receives it is when a server’s admins connect it themselves, as described below.
Memory search. So the AI can find the right entry in your server’s brain, NEXUS keeps a search index of it, built by Voyage AI. Voyage AI receives the text of the entries your staff taught and the search terms the AI writes when it looks something up, never the ticket conversation itself. This is part of how the AI answers on every server and cannot be switched off on its own.
Web lookups. When an answer needs current information from outside your server, the AI can run a web search through Brave Search. Only the search terms the AI writes are sent, never the ticket conversation. This is on by default and a server admin can switch it off for the whole server in the bot’s /ai settings.
If your server connects its own AI
On plans that include it (Pro, and Enterprise where the deal does), an admin with the Manage Server permission can connect the server’s own account at an AI provider of their choosing, for example OpenAI, Anthropic, Google, OpenRouter, DeepSeek, xAI, Mistral or Groq, by entering its address, an API key and a model name. While it is in use, the AI calls for that server go to that provider instead of to Anthropic: ticket replies, the public FAQ channel, the website widget, the dashboard chat and training. Each call carries what the model needs to answer: the conversation, the relevant brain entries and our instructions to the model.
That provider is chosen by the server, not by us, and processes the data under the server’s own agreement with it. It is not a sub-processor we engage, so it is not listed in section 4, and its retention and training practices are its own: check them before connecting. Memory search and web lookups keep using Voyage AI and Brave as described above. If the provider rejects the key, runs out of credits or stops offering the model, the server switches back to our AI, and from then on Anthropic handles its tickets as described above; the admin who connected it is told by direct message.
The API key. It is sent once, over an encrypted connection, to NEXUS, where it is stored encrypted and used only to make that server’s AI calls. It is never shown again, apart from its last four characters, and this website never stores or logs it. Connecting, changing the model and testing each make one small real call on the provider account. We keep a record of how many tokens each call used, so the dashboard can show usage, but we do not charge for those replies. Removing the provider deletes the key. If the server’s plan stops including the feature, the key is kept but switched off, so it can resume on renewal; an admin can remove it at any time.
4. Who else handles your data (sub-processors)
We rely on a small set of trusted sub-processors to operate the service. Each is listed below with its role, region, and own privacy policy. If we add a new sub-processor that processes user content, we will update this list before the change takes effect.
| Provider | Role | Region |
|---|---|---|
| Anthropic | AI model inference (via NEXUS) for every server that has not connected its own AI provider. Processes message contents to generate responses. | United States |
| Voyage AI | Search indexing for each server's AI memory (via NEXUS), so the AI can find the right entry. Receives the text of the entries a server's staff taught and the search terms the AI writes, never the ticket conversation. Our account is opted out of Voyage AI using this content for training, so it is deleted right after it is processed. | United States |
| Hetzner Online GmbH | Hosting and database storage for the bot, dashboard, and NEXUS infrastructure. | European Union (Germany / Finland) |
| Cloudflare | CDN, DDoS protection, and DNS for the marketing site and API. | Global (EU edge nodes serve EU traffic) |
| Stripe | Card and bank-transfer payments for plan subscriptions and token packs. Under Stripe Managed Payments, Stripe is the merchant of record: it takes the payment, collects and remits any VAT or sales tax, may charge in your local currency, and issues your receipt. | European Union (Ireland) for EU customers |
| NOWPayments | Cryptocurrency payment processing for plan subscriptions and token packs. | European Union |
| Brave Search | Live web lookups when the AI needs current information from outside your server. Receives only the search terms the AI writes, never the ticket conversation. Server admins can switch this off per server. | United States |
| Discord | Identity provider via OAuth. Bot platform. Payment processor and merchant of record for subscriptions bought inside Discord, where Discord takes the payment and handles cancellation and refunds itself. We receive only your public Discord ID, username, avatar, and the list of servers where you have admin rights. | United States |
| Google Analytics | Aggregate marketing-site traffic analytics. Only loaded if you grant Analytics consent. | European Union (Google Ireland) |
Where data is transferred outside the EEA (notably to Anthropic, Voyage AI, Brave and Discord, all in the United States), the transfer is covered by the EU Standard Contractual Clauses, or an equivalent safeguard, in those providers’ data-processing terms. An AI provider a server connects itself is covered by that server’s own agreement with it.
5. How long we keep data
We keep data only as long as we need it to operate the service or to comply with a legal obligation.
| Data | Free | Premium | Pro | Enterprise |
|---|---|---|---|---|
| Ticket transcripts | 90 days | 730 days | 730 days | 730 days |
- Bot removal from a server: when the bot is removed from a Discord server, all data associated with that server (panels, transcripts, brain content, blacklist entries, audit log rows tied to that server) is permanently deleted within 48 hours. The 48-hour window exists so an accidental kick can be reversed if you re-add the bot, but is not a retention promise: once the window closes, the data is gone.
- Account data: kept while your account exists. On account deletion, your identifying records are removed within 48 hours; see “Your rights” below.
- Audit logs: retained for up to 12 months for accountability, then purged. The dashboard’s “all time” audit view therefore shows at most this window.
- Ticket feedback ratings: the 1 to 5 star ratings and notes members leave when a ticket closes are retained for up to 24 months, then purged.
- Statistics and staff activity: the per-staff contribution data behind the dashboard’s statistics and leaderboard is retained for up to 24 months, then purged. Where the dashboard offers an “all time” range, it reflects at most the data still within these retention windows, not the entire history of the server.
- Billing records: retained for the period our tax obligations require (typically 5 years in Romania) for audit and accounting. Anonymized after the retention period ends.
- Marketing analytics: per Google Analytics 4 defaults, up to 13 months.
6. Your rights
Under the GDPR and equivalent laws, you have the right to:
- Access the personal data we hold about you.
- Have inaccurate or incomplete data corrected.
- Have your data deleted (the “right to be forgotten”).
- Restrict or object to certain processing.
- Export your data in a structured, portable format.
- Withdraw consent for analytics at any time through the cookie banner.
- Lodge a complaint with a supervisory authority. In Romania this is ANSPDCP; in any EU country you may use your local authority.
Logged-in users can self-serve export and deletion from Account settings in the dashboard. Both run immediately: the export includes your complete AI conversation history, and deletion permanently erases it (your id becomes a pseudonym and message content becomes unreadable) along with everything this site holds about you. Records we are legally required to keep (billing, up to 5 years) or need to prevent reward abuse (your vote and reward history, so an already-claimed reward cannot be claimed again) are retained. One honest caveat: facts a server’s staff taught their bot’s AI brain are that server’s knowledge base, not a per-user record; if one contains your personal data, ask the server’s staff to remove it or email us. For anything else the self-serve flow does not cover, email [email protected] from the address on file or from a Discord-verifiable channel and we will respond within 30 days, as required by law.
7. Cookies
We use cookies to make the site work and, only with your permission, to understand which pages help people the most. Full details, including a per-cookie table, live on the cookie policy page. You can change your choice at any time from that page.
8. Security
We protect data with HTTPS in transit, encrypted storage for sensitive configuration (webhook URLs, OAuth tokens), strict access controls on production infrastructure, and rate-limited API endpoints. No method of transmission or storage is 100 % secure; if we ever become aware of a breach affecting your personal data, we will notify affected users without undue delay and inform supervisory authorities as required by Article 33 of the GDPR.
9. Children
AI Ticket Bot is not directed at children under 13. Discord’s own Terms of Service require users to be at least 13 years old (or older where local law requires). If we learn we have collected data from a child under 13, we delete it.
10. Changes to this policy
We update this policy when the service changes in a way that affects how data is handled. The effective date at the top of this page reflects the last update. Material changes are surfaced via a dashboard notification and a re-prompt of the cookie banner where relevant. Continued use after the effective date constitutes acceptance.
11. Contact
For privacy questions, data-rights requests, or anything related to this policy:
- [email protected] for support and data-rights requests
- [email protected] for legal notices and business matters