Privacy Policy
Effective date: 18 November 2026
1. Who we are
Beegent (beegent.ai) is operated by Beegent.ai (“Beegent”, “we”, “us”). We are the controller of the personal data described here.
Contact for anything about privacy: hey@beegent.ai.
2. What Beegent does
Beegent reads the sources you connect (for example Telegram, Discord, Substack, X, YouTube, RSS feeds, forwarded newsletters, or content you send through our API), summarizes them with AI, and shows or sends you briefings, signals and search results.
3. What we collect and why
| Data | Why we use it |
|---|---|
| Account: email address, optional name, password (held by our sign-in provider, Amazon Cognito; we never see it in plain text) | Create and secure your account, send verification and account emails |
| Subscription: plan, status, trial and renewal dates, Stripe customer and subscription IDs. We do not receive or store your full card number | Provide the plan, enforce its limits, handle billing |
| Preferences: time zone, interests, your goal in your own words, sensitivity, briefing schedule, signal rules and followed names, feedback verdicts (Useful, Not relevant, Incorrect) and any note you add to an Incorrect verdict | Personalize and rank your briefings |
| Connections: which sources you connected and the names you gave them | Fetch and show your sources |
| Credentials for sources you connect: a Discord token, a Telegram session, Substack cookies, or a Discord webhook URL for delivery. These are stored encrypted (see section 10) | Read your sources and deliver to your Discord channel on your behalf |
| Source content: messages, posts, articles, video transcripts and newsletters from the sources you connect or send us, including authors’ names and handles, timestamps and links. Forwarded newsletters include the sender’s name and address; we do not store your own receiving address | Store, summarize, search and show you what happened |
| Derived content: summaries, developments (events), ticker and sentiment data, briefings, signal matches, and a search index built from the above | The core service |
| Delivery settings: your email address, a Telegram chat ID and title (groups only), a Discord channel label, send history markers | Send briefings and alerts where you asked |
| API keys: only a one-way hash of each key, its name, scope and last-used time | Authenticate your API calls |
Ad measurement, only if you accept cookies: the ad click IDs and campaign tags in the link that brought you to Beegent (for example gclid, fbclid or utm_source), the page you landed on, the site that sent you, and the IDs in the Google and Meta cookies listed in section 11. When you start a subscription, also a hashed (SHA-256) copy of your email address, your IP address and your browser’s user agent |
Measure which of our ads lead to sign-ups and subscriptions (see section 11) |
| Technical and usage records: request logs with a hashed user reference, counts and error codes. Our AI usage records contain token counts and model names only and never contain prompt text, message text, user IDs or emails | Operate, secure, monitor costs and fix bugs |
We do not sell your personal data.
4. Sources you connect
You decide what to connect. You are responsible for having the right to connect each source and to have us process its content, including content that belongs to other people (for example messages in a group or server). Do not connect sources you are not allowed to access or share. See our Terms.
Some sources are shared across customers to save work. If several customers follow the same public channel, we may fetch it once and reuse the stored content and summaries for those customers, each of whom must have their own connection to it. Private groups of X accounts or YouTube channels you build, and your forwarded-newsletter inbox, are used only for your account.
Telegram accounts. When a Telegram channel cannot be read by the account of one of the customers connected to it, we may read it with another account connected to Beegent that is still allowed in that channel, for example an account you connected. We only read channels that account is already allowed in, we check this with a read-only lookup, and we never join, leave or post. If you connect a Telegram account, you agree it may be used this way for channels it can already access.
5. AI processing
To produce summaries, briefings, signals and search, we send the relevant source content and, where needed, your stated goal or interests to large language models on Amazon Bedrock (AWS), and we create text embeddings for search with an Amazon Titan model. The models are run by AWS on our behalf. Per AWS documentation, model providers do not have access to Bedrock logs or to customer prompts and completions.
We do not use your data to train AI models. Summaries are generated automatically, can be wrong, and link back to source messages so you can check them.
6. Who processes data for us
| Provider | Purpose | Location |
|---|---|---|
| Amazon Web Services | Hosting, database (DynamoDB), search index (Aurora PostgreSQL), message queues, encryption keys (KMS), file storage for inbound mail, logs | United States (us-east-1, N. Virginia) |
| Amazon Bedrock (AWS) | AI summaries, briefings, signal matching, embeddings | United States. Some models are called through AWS inference profiles that may process a request in another US region |
| Amazon Cognito (AWS) | Sign-up, sign-in and account email verification | United States |
| Amazon SES (AWS) | Sending briefing emails and receiving forwarded newsletters | United States |
| Stripe | Payments, subscriptions, invoices, tax calculation, and holding the ad measurement details of a checkout (section 11) | United States and other countries where Stripe operates |
| RapidAPI and the data providers behind it | Reading X posts and YouTube captions and playlists. We send account handles, list or playlist or video identifiers, not your personal data | Provider dependent |
| Telegram | Reading Telegram channels with connected accounts; sending briefings through our Telegram bot | Telegram’s own infrastructure |
| Discord | Reading Discord channels with a connected token; sending briefings to your webhook | Discord’s own infrastructure |
| Substack | Reading Substack posts with your connected cookies | Substack’s own infrastructure |
| Google (Google Tag Manager, Google Analytics, Google Ads, Google Fonts) | Loading Google Tag Manager on beegent.ai and in the web app, and fonts in the web app. Only if you accept cookies: analytics and measuring which Google ads lead to sign-ups and subscriptions (see section 11) | United States and other countries |
| Meta Platforms (Meta Pixel, Conversions API) | Only if you accept cookies: measuring which Facebook and Instagram ads lead to sign-ups and subscriptions (see section 11) | United States, Ireland and other countries |
We choose these providers to act on our instructions. We will update this list when it changes. Google and Meta also use the ad measurement data they receive under their own terms, for example to match a conversion to a Google or Meta account you have; for that use their own privacy policies apply.
7. International transfers
Our systems run in the United States. If you are in the EEA, UK or Switzerland, your data is transferred to the United States. For these transfers we rely on the safeguards our providers offer, such as the European Commission’s Standard Contractual Clauses in their data processing terms.
8. Legal bases (EEA and UK)
We process data to perform our contract with you (providing the service), for our legitimate interests (security, abuse prevention, improving quality, cost control), to comply with the law (tax and accounting), and with your consent where required, for example optional emails you turn on, and analytics and advertising cookies and ad conversion measurement (section 11). You can withdraw consent at any time.
9. How long we keep data
- Account and preferences: until you delete your account.
- Source messages, summaries, events and search index: kept while the service runs and not automatically expired, except as listed below. Content from a shared channel may remain after you leave if other customers still follow it, but it is no longer linked to you.
- Hourly briefings: about 30 days after their window ends. Daily, weekly and monthly briefings are kept until account deletion.
- Quota and rate-limit counters: days to a few weeks.
- Ingest item IDs (used to detect repeats): 30 days.
- Browser login sessions for connecting Discord or Substack: minutes; the captured credential is cleared after you apply it.
- Forwarded raw emails: deleted right after processing, and at the latest after 3 days.
- Application logs: 14 to 30 days.
- Database backups: point-in-time recovery is on, so deleted data may remain in backups for up to 35 days.
- AI usage records: kept without a time limit, contain no personal content.
- Billing records: held by Stripe and by us as long as tax and accounting law requires.
- Cookie choice (
bg_consent): 180 days in your browser; after that we ask again. - Ad attribution cookie (
bg_attr): 90 days in your browser. A later visit from another ad replaces it, and choosing Reject deletes it. - Ad measurement details of a subscription: kept with the checkout and subscription records at Stripe for as long as those records exist. Google and Meta keep what we send them under their own retention rules.
If your subscription ends (cancelled, unpaid or expired): we pause your connections and the Telegram, Discord and Substack accounts you connected so nothing is read or processed for you. We keep your stored credentials and data so that resubscribing works. If you want them removed, delete your account.
If you delete your account: we remove your connections, your connected Telegram, Discord and Substack accounts and their credentials, browser sessions, pending logins, your briefings, your preferences, delivery settings, feedback, signal records and your sign-in identity. Not part of this routine: your API key records (we keep only a one-way hash of each key), messages you sent us through the API, and content of channels that other customers still follow. Email hey@beegent.ai and we will delete the first two as well. We also delete your customer record at Stripe, but Stripe and we may keep invoices and payment records we are legally required to keep.
10. Security
- Traffic is encrypted in transit (HTTPS).
- Discord tokens, Telegram sessions and API hashes, Substack cookies and Discord webhook URLs are encrypted before storage with keys in AWS KMS. The API does not return them to you or anyone else.
- API keys are stored only as hashes and shown once.
- Passwords are handled by Amazon Cognito.
- Access to stored data is limited through AWS IAM roles.
- Sensitive values are kept out of our logs.
No system is perfectly secure. Our own staff and systems can technically access data needed to run the service, and a compromise of our running service could expose credentials in use. If we learn of a breach affecting you, we will notify you and authorities as the law requires.
11. Cookies, analytics and advertising
Your choice. The first time you visit beegent.ai or the web app, a banner asks whether you accept cookies that measure our ads and help us improve Beegent. Analytics and advertising cookies are set only after you choose Accept. Reject is as easy as Accept, and Beegent works the same either way. Your choice is stored on beegent.ai and its subdomains, so it applies to both the site and the web app.
| Cookie | Set by | Purpose | Lifetime |
|---|---|---|---|
bg_consent |
Beegent | Remembers whether you accepted or rejected | 180 days |
bg_attr |
Beegent, only after Accept | Remembers the ad or campaign link that brought you: its click IDs (such as gclid or fbclid) and utm_ tags, the page you landed on, the site that sent you and the time, so we can credit the right ad if you subscribe |
90 days |
_ga, _ga_* |
Google Analytics, only after Accept | Count visits and show how people use the site and the app | Up to 2 years |
_gcl_* |
Google Ads, only after Accept | Connect a sign-up or subscription to the Google ad you clicked | 90 days |
_fbp, _fbc |
Meta Pixel, only after Accept | Connect a sign-up or subscription to the Facebook or Instagram ad you clicked | 90 days |
Google Tag Manager. Every page of the site and the app loads Google Tag Manager, a Google script that loads the other tags we choose. When it loads, Google receives your IP address and browser details. Until you accept, Google’s tags run in a restricted mode that sets no cookies, although Google may still receive signals without cookies, such as that a page was viewed. Meta’s tags run only after you accept.
Ad conversion measurement. If you accept cookies, the Google and Meta tags in your browser record the steps of signing up and subscribing (for example sign-up, checkout and the start of a trial) together with the cookies above. When you then start a subscription, including a free trial, our servers also report it directly, so the count is right even when a browser blocks the tags:
- to Meta: a hashed (SHA-256) copy of your email address and of your Beegent account ID, your IP address, your browser’s user agent, the Meta cookie IDs (
_fbp,_fbc, or thefbclidfrom your ad link), the page address, and the plan, price and currency; - to Google: your Google Analytics cookie ID, the order ID, and the plan, price and currency.
We send the same report when a free trial turns into a paid subscription. Hashing hides your email address in transit, but Meta can still match it to a Meta account that uses the same address. To connect these details to your subscription, we store them with your checkout and subscription records at Stripe. If you rejected cookies, our servers send nothing.
Changing your mind. Use the Cookie settings link at the bottom of any page on beegent.ai to change your choice at any time. Choosing Reject stops the tags from setting cookies and deletes bg_attr and the Google and Meta cookies listed above wherever our site can reach them. It does not recall reports already sent. A subscription you started while cookies were accepted keeps that choice for the report when its trial turns into a paid plan; email hey@beegent.ai if you want us to stop it.
Local storage in the web app. The app uses your browser’s local storage to keep you signed in (sign-in tokens and basic profile) and to remember small display choices such as your briefing view mode and channel names, and session storage to remember the plan you picked while you are at checkout. These are strictly necessary or functional. The app also loads fonts from Google Fonts, which gives Google your IP address and browser details.
12. Your rights
Depending on where you live, you may have the right to access, correct, delete, export and restrict or object to the processing of your personal data, and to complain to your data protection authority. You can edit most data and delete your account in the app. For anything else, or to request an export, email hey@beegent.ai. We will respond within the time the law requires and may need to verify your identity.
EEA, UK and Switzerland (GDPR). The rights above apply, and you may withdraw consent at any time.
California (CCPA/CPRA) and similar US state laws. You may request to know, access, correct and delete your personal information, and not to be discriminated against for exercising these rights. We do not sell personal information. Giving Google and Meta the ad click and cookie IDs described in section 11 may count as “sharing” for cross-context behavioral advertising under California law. It happens only after you accept cookies, and you can opt out at any time with Cookie settings or by emailing hey@beegent.ai.
13. Children
Beegent is not for anyone under 16 (or the minimum age in your country). We do not knowingly collect data from children. If you believe a child has given us data, contact hey@beegent.ai and we will delete it.
14. Changes
We may update this policy. For material changes we will notify you by email or in the app before they take effect. The effective date above shows the latest version.
15. Contact
Beegent.ai. Email: hey@beegent.ai.