Uxen Privacy Policy
1. Who we are
Uxen is a personal assistant for your own correspondence. The service is operated by Ilya Romanovich Romanov.
For any question about your data, write to privacy@rrrfirstchat.com. We reply from this address, and it is the address for every request described in this policy.
We do not publish a phone number here on purpose: public pages are crawled by contact scrapers, and a published number ends up in spam databases. Please use the address above.
2. Summary
Uxen answers questions about correspondence you connected yourself: your mail, Telegram, WhatsApp, your calendar. Nothing is connected without an action by you, and any source can be disconnected in one step.
To answer, Uxen reads that correspondence, and to make it searchable, its text is sent to our AI model provider. Section 5 describes this in detail.
Uxen can write to people on your behalf — only where you have switched that on. By default it writes to no one: the autoreply and “write for me” drafts stay off until you turn them on. Since 17 August 2026 this also covers email: a separate button lets you allow sending letters on your behalf and sorting mail into folders (mark read, archive, apply a label). Uxen cannot delete email. A draft is never sent without your confirmation. Everything sent on your behalf is recorded, and you can review it.
3. What we collect
To sign you in. Your email address, and, if you sign in with Apple or Google, the account identifier and email address they provide. There is no password: email sign-in uses a one-time code. The code itself is not stored — only its cryptographic hash, and it can be used once.
To run the app. An account identifier, a device identifier, the device name, platform and app version, and, if you allow notifications, a push token.
To protect sign-ins. The approximate location of each signed-in device: city and country. The server derives them from the device's last IP address using the IP to City Lite database by DB-IP, hosted on our own server. The device record does not keep the address itself; web-server logs keep it for 14 days. The location appears in your device list and in the “New sign-in” email so that you can spot a sign-in that isn't yours. We never request or receive your phone's location.
To investigate failures. For each assistant answer we keep a technical record for 30 days: how the question was understood, where information was looked for, and what the answer was based on. The record is encrypted with your key, and Uxen staff open it only under the rules in section 10. If you tap “Report a problem” under an answer, we keep your feedback for 30 days. If you connected Claude, ChatGPT or another assistant, we keep the list of its requests to your memory for 90 days so you can review it in the Security section.
Account history. Records of important account events: when a sign-in method was linked, confirmed, declined or unlinked, and when the same email address appeared on two accounts. The records say what happened and when. They never contain passwords, codes or tokens.
Content of the sources you connect. Message and email text, subject lines, sender and recipient names and addresses, dates, chat names, calendar events and, if you connect them, your Google contacts. Without this data the service cannot answer questions about your correspondence.
Attachments: documents. From documents — files, spreadsheets, presentations — that arrive in a connected conversation or that you attach to a question, the service extracts the text on our server, without passing the file to third parties, and stores only that text. Voice messages, photos and videos are not transcribed or recognised.
Attachments: a service reference instead of the file. We do not store the files themselves from WhatsApp. Next to the copy of the message we keep only what WhatsApp sent about the attachment: its type, size, name, and a service reference to it together with the decryption key, without which WhatsApp will not release the attachment. The service uses this reference to fetch a document when it extracts its text. The reference is deleted together with the copy of the conversation.
Calls. Since 16 August 2026 the service receives information about calls from the messengers you connect: with whom, when, incoming or outgoing, answered or missed, duration, and whether it was a video call. This makes it possible to answer questions such as “when did I last speak to this person”.
Call content is not recorded. The service does not listen to calls, does not join them and keeps no audio: the messengers provide only the information listed above. These records are kept as long as the correspondence and are deleted the same ways: “Erase this source's data”, “Erase all data”, and account deletion.
Files are not kept. The service stores the extracted text, not the file. During extraction the file is held in process memory or in a temporary file readable only by our service, and is deleted right after reading. A temporary file left behind after a failure is removed within an hour.
How long text from files is kept. Text from a file you attached to a question becomes part of your memory and is kept while you use the service. Deleting a conversation with the assistant does not remove it; “Erase all data” and account deletion do.
If you picked a file but never asked about it, its text is deleted within a day. Until then it is kept in a service record next to your in-app conversations.
Information about the people you correspond with. To recognise the same person across messengers and to answer “who is this”, the service stores what the messenger reports about your contact and what the person published about themselves: their name and display name, username, phone number, email address, birthday from their profile, “about” text, the number of their profile picture (for Telegram, also a small blurred copy of it) and, from Google Contacts, their organisation and job title if you recorded them. If you asked to be told when someone comes online or starts typing, the service watches that status while your request is active.
The same person is linked across messengers automatically only by phone number; a matching name only leads to a suggestion that you link the profiles. Of a profile picture we keep only its number, and for Telegram also a blurred copy of about 1 KB, which Telegram sends together with the rest of the profile. WhatsApp reports the picture's number when it changes or in answer to our request — no more than once a week per conversation; the contact is not notified of this request. The sharp picture is fetched from the messenger when it is shown to you and is not stored on the server; your device keeps it in its cache for up to a week. Everything else listed arrives in the responses the messenger already sends while the service reads your correspondence; no separate requests are made for it. The service treats an “about” text as the contact's own words, not as a fact or an instruction.
Your conversations with the assistant. Your questions and the assistant's answers.
Support requests. If you write to us through “Support” on the website, in the app or in the bot, we receive the text of the request and up to three attached files. They are used only to respond and to fix problems; the reply goes to the email address verified when you signed in. Requests and files are kept for 90 days and deleted at once if you delete your account.
To connect a source. For Telegram, the phone number, the login code and, if two-factor authentication is on, your Telegram cloud password. These are passed directly to Telegram to complete the sign-in, are not stored in our database and are not written to logs.
For Instagram and Messenger, the username, password and confirmation code. These are passed directly to Meta and are not stored anywhere on our side. Uxen only reads your direct messages: it sends nothing and marks nothing as read. The sign-in uses an unofficial method, so Meta may ask you to confirm the sign-in or temporarily restrict the account.
For mail without Google's permission, and for your studies. An IMAP mailbox, Gmail with an app password, the student mail and Nomotex connect with a login and a password: they have no official permission like Google's. Yandex Mail connects with Yandex ID or an app password, Yandex Calendar with an app password. The password is stored only in encrypted form (AES-256), is used to read on your behalf, never goes into logs or backups, and is deleted when you disconnect the source or erase your data. The university account password is not stored — only the session is kept after sign-in.
When you pay. The amount, the date and the payment reference. Card details never reach us: they are entered on the payment provider's side and do not pass through our servers.
Technical logs. Identifiers, chat names, counters and operation durations — information needed to find faults. They contain no message text.
Cookies and browser storage on the website. The site uses one cookie — a sign-in key valid for 90 days. Your browser sends it only to our server, page scripts cannot read it, and “Sign out” deletes it. Browser storage holds only your screen and language settings, a sign-in marker, your account number and sign-in method (none of which give access to your account), the numbers of events you were already notified about, marks on dismissed “is this the same person?” prompts, and a copy of the site's pages for faster loading. There are no analytics, advertising or third-party cookies, so the site does not ask for cookie consent.
4. What we use it for — the complete list
The data above is used only to:
- answer your questions about your correspondence;
- maintain the memory in which answers are found;
- find commitments and tasks in your correspondence and remind you of them;
- if you switch it on, reply and write on your behalf, sort your mail, add agreements to your calendar, and tell you when a person you named comes online;
- if you switch it on, keep you “online” in Telegram or WhatsApp for the period you choose, and show “typing” before an autoreply;
- sign you in on your devices;
- protect your account and stop abuse;
- respond to support requests;
- send service messages: a sign-in code, a warning before deletion, a notice that a connection has broken;
- take payment and manage your subscription;
- find and fix faults and monitor how the service works;
- comply with orders from authorities that are binding on us, under the rules of section 11.
There are no other purposes. We do not use your data for advertising, do not build advertising or behavioural profiles, do not sell your data, do not share it with data brokers, and do not combine it with data from other services to track you. There is no advertising in the product.
5. Who else receives your data
All recipients of your data are listed below.
OpenAI — the AI models. OpenAI receives data in two cases:
- when you ask a question — the question and the related fragments of correspondence, to compose the answer;
- for search across your correspondence — the text of everything Uxen indexes, from which numerical representations (embeddings) are computed. This happens in the background, not only when you ask.
As a result, the content of your correspondence leaves our server and is processed by OpenAI. We use the OpenAI API under terms that prohibit training OpenAI's models on this data. OpenAI is based in the USA and, under its API terms, keeps request logs for up to 30 days to prevent abuse.
This sharing starts only after your explicit consent. The service asks for it on a separate screen right after the first sign-in — before any source is connected and before the first question. Without consent nothing is sent to OpenAI, but the service cannot answer questions either; the same screen lets you sign out or delete the account.
Google — only if you connect Gmail, Calendar or Contacts. The service calls Google's APIs on your behalf under the permission you granted. In all three cases the basic permission is read-only: without write permissions the service changes nothing and sends nothing. To identify the mailbox owner and your name, the service receives your Google account's email address and name.
Each write permission is granted separately — with its own checkbox on Google's consent screen. When you connect, Google lists the permissions and you can untick any of them; later a permission is granted with its own “Allow” button. Until a permission is granted, the corresponding action is not performed. You can withdraw a permission in your Google account settings (myaccount.google.com/permissions) or by disconnecting the source in Uxen.
- Writing events to your Calendar. An agreement you confirm is added to your calendar: the event title, the place, the chat name and a link to the original message, plus a description and participants if you provided them; Google will send the participants an invitation. The conversation itself is not copied. If you cancel the agreement, the event is cancelled. Events not created by Uxen are not changed.
- Sending mail on your behalf. A letter is sent only after you confirm the finished text. The narrowest available permission is requested — sending only. It does not allow reading, changing or deleting mail.
- Sorting your mail into folders. Mark read, archive, apply and create a label. This Google permission also allows moving mail to the bin, but Uxen does not do this — the software has no such function. The permission does not allow permanent deletion.
Contacts are read-only. The service cannot change your contacts: we do not request that permission.
Google API Services User Data Policy — Limited Use. Uxen's use of information received from Google APIs adheres to the Google API Services User Data Policy, including the Limited Use requirements. Data from Gmail, Calendar and Contacts is used only for the features you asked for. It is not sold, not used for advertising, and not transferred to third parties except as needed to provide those features (see the model provider above), to comply with the law, or as part of a merger or acquisition — and then only with your prior consent. Humans do not read it unless you explicitly allow it, to fix a problem you reported, for security reasons, or where the law requires it. Google user data is not used to develop, improve or train generalised AI models: the model provider processes it under API terms that prohibit training on this data.
Resend (USA) — our email provider. All our emails to you are sent through it: your address, the subject and the body, including the one-time sign-in code.
An assistant you connected yourself (Claude, ChatGPT or another). The assistant receives data only after you connect it, and only in response to its own request: messages and emails found, a whole conversation, calendar events, study information, the list of your conversations, and from Telegram also attachments. From then on, that data is processed by the assistant's company under the terms of your account there. You can disconnect the assistant in Security, where all its requests are visible for 90 days.
Apple. If you use Sign in with Apple, Apple confirms your identity and gives us an account identifier and an email address. If notifications are on, Apple's push service receives your device token and the short text of the notification — for example, a contact's name, the title of a task found or the beginning of a finished answer; it never receives your correspondence as a whole. In a browser, notifications are delivered by the browser's own push service (Google, Mozilla or Apple); their content is encrypted and that service cannot read it.
Telegram. If you connect Telegram as a source, our server acts as a linked device on your account through Telegram's official interface — you can see it in your device list and disconnect it there. If you use the Uxen bot, your messages to the bot and its replies pass through Telegram. Payments in Stars are processed by Telegram. We read support requests in the service section of the Uxen bot, so their text and screenshots pass through Telegram.
WhatsApp (Meta). If you connect WhatsApp, our server acts as a linked device on your account — you can see it in WhatsApp's linked-devices list and disconnect it there. If you connect WhatsApp Business, Meta sends us the business number's messages using the keys you entered; the keys are stored encrypted.
Instagram and Messenger (Meta) — only if you connect them. Our server signs in to your account and reads your direct messages; to Meta this looks like an ordinary sign-in.
Yandex (Russia) — only if you connect Yandex Mail or Calendar. The service accesses your mailbox and calendar on your behalf, the same way any mail program does.
YooKassa — only when you pay by card or SBP. YooKassa receives the amount, the currency, a payment description and your account identifier. Card details are entered on YooKassa's side and never reach us. No correspondence data is sent with a payment: the payment system and the service's memory are not connected.
Aeza — hosting. The server that stores and processes your data is in the Netherlands (Amsterdam).
Anthropic (USA) — our team's AI assistant for development and failure investigation. It receives what a specific task needs; email addresses and phone numbers are replaced with placeholders, but message text and names may be passed to it.
No one else. Your data is not shared with advertising networks, data brokers or analytics services. There are no tracking scripts or advertising libraries in the product.
What we require of them. The providers that process data on our behalf — OpenAI, Resend, Aeza, Anthropic, YooKassa — work under terms that prohibit them from using our data for their own purposes and oblige them to protect it; each receives only what its service needs. Services you connect yourself (Google, Telegram, WhatsApp, Instagram, Messenger, Yandex, your assistant) hold your data under your own agreement with them.
6. What Uxen does not do
- it writes to no one on your behalf until you switch that on — the function is off by default, and a draft is never sent without your confirmation;
- it does not delete your email — the software has no such function;
- it does not track you across other apps and websites;
- it shows no advertising and builds no advertising profiles;
- it does not read sources you have not connected;
- it does not ask for your Google mail password when you connect it with the Google button. Gmail can also be connected with an app password — then it is stored encrypted, like the passwords of other mailboxes (IMAP, the student mail, Yandex); see “For mail without Google's permission, and for your studies”.
Connecting Telegram may require your two-factor cloud password: Telegram requires it to add a device. The password is passed directly to Telegram to complete the sign-in and is not stored or logged. If you prefer not to enter it, do not connect Telegram — everything else works without it.
7. How long we keep things
Your memory is kept while you use Uxen and is deleted when you delete it (section 8). Uxen is built to answer questions about years of correspondence, so memory is not deleted automatically after a set period: otherwise questions about earlier periods could not be answered.
Other data has these limits:
- technical logs (identifiers and chat names, no message text) — 14 days;
- system logs on the server — 14 days;
- sign-in tokens — 90 days; they are revoked immediately when you sign out;
- one-time sign-in codes are valid for a few minutes, stored only as a hash, and usable once;
- the technical record of an answer and “Report a problem” feedback — 30 days; requests from connected assistants to your memory — 90 days; our internal log of staff access to your data — for as long as the account exists;
- support requests and their attached files — 90 days;
- backups are made daily. The server keeps the last 7 daily backups and one weekly; the copy held by the operator is stored only encrypted — 14 daily and 3 weekly. Deleted data disappears from backups as they are replaced — within 3 weeks.
Inactive accounts. Accounts are not currently deleted automatically. If we introduce automatic deletion of long-inactive accounts, we will give at least 30 days' notice — in the app, by email, and in the bot if you use it — and any question to the assistant will extend the period.
If Uxen shuts down or is transferred. If Uxen shuts down, we will notify you in advance, let you export your data, and delete it. If Uxen passes to another company, we will notify you in advance, and your data will move to it only with your consent; before the transfer you can export and delete it.
8. Deleting your data
Uxen has four different actions, with different consequences.
Disconnect a source. New messages from the source stop arriving. Everything already stored stays and is used in answers: disconnecting stops collection but does not erase memory. You can reconnect the source at any time.
Erase a source's data. Everything Uxen stored from one source is deleted; your other sources and the account remain. This cannot be undone. If the source stays connected, Uxen keeps receiving new messages from that moment on. Profile pictures of your contacts stay until you disconnect the source, erase all data or delete the account.
Erase all data. All sources are disconnected, everything stored is deleted, and your conversations with the assistant are deleted. Your account and sign-in methods remain — you can connect sources again. This cannot be undone and requires confirming your identity.
Delete the account. Everything “Erase all data” does, plus the account itself: sign-in methods, verified email addresses, devices, push tokens and the account history. Requires confirming your identity. Deletion takes effect after 7 days. Right after you confirm, your sources are disconnected, you are signed out on all devices, connected assistants lose access, and an email is sent to your verified address. If you sign in within those 7 days and tap “Cancel deletion”, the account and stored data remain, but you will need to reconnect your sources. After 7 days the deletion is final. “Erase all data” takes effect immediately, with no delay.
What remains after deletion. Two records:
- your plan and its paid-through date. If you delete your account before the paid period ends and come back later, the remaining days are restored;
- the payment ledger — the reference, amount, status and date of each payment. It is needed to reconcile payments, handle refunds and meet tax requirements.
Neither record contains the content of your correspondence.
You can also remove Uxen in the messenger itself, in its linked-devices list.
9. Your choices, and how to withdraw consent
Uxen works with a source only under your permission, and any permission can be withdrawn:
- disconnect any source — in the app, in one step;
- revoke Google's permission — in your Google account, at myaccount.google.com/permissions; this takes effect regardless of us;
- remove the Uxen device in Telegram or WhatsApp — in their linked-devices list;
- turn off notifications — in your device settings;
- erase all data or delete your account, as described in section 8. Deleting your account is a complete withdrawal of consent: nothing is left to process.
You can download a copy of what Uxen stored about you — conversations with the assistant, memory, people, tasks, cards, sign-ins and the security log — at any time: Settings → Security → “Download my data”. The correspondence itself from messengers and mail is not included — it stays with you in those services. Any other data we hold about you we will send on request to privacy@rrrfirstchat.com. If you find an error in your data, write to the same address.
10. Security
Data travels only over encrypted connections. Conversations with the assistant and what Uxen extracted from your correspondence are stored encrypted on the server, with a separate key for each user. Sign-ins to your accounts are stored the same way: Telegram sessions, Google access, mail and Nomotex passwords, the university account session. Copies of messages used for search, WhatsApp sign-ins, WhatsApp Business messages and support requests are stored without separate encryption: they are protected by restricted access to the server, not by your key. To search and answer, the server decrypts data, so this is not end-to-end encryption.
To fix failures and respond to requests, authorised Uxen staff may view the data related to a specific problem. Beyond that, data is opened only to investigate an attack or abuse, or under a binding order from an authority, under the rules of section 11. Only those who need access for this work have it, and every access is recorded in our internal log. Data received from Google is subject to the stricter rule in section 5 (Limited Use): staff view it only with your permission, to fix a problem you reported, for security reasons, or where the law requires it. A sign-in from a new device triggers a notification to your other devices and, if an email is linked to the account, an email.
Access tokens are kept in the device's secure storage (Keychain), not in ordinary app settings. On the server, key files and databases are readable only by the server administrator. Sign-in codes and sign-in tokens are stored as hashes; reuse of a token revokes the whole chain of tokens linked to it. Backups are made daily and checked for integrity; one copy is kept away from the main server, encrypted only. More details: rrrfirstchat.com/security.
No system is completely secure. If a breach affects your data, we will tell you.
11. Requests from government authorities
- We do not hand over your data voluntarily — not to government authorities, not to companies, not on informal requests.
- We disclose data only under an order that is legally binding on us under the law of the Netherlands, where the data is stored, or of the European Union. We check every order for lawfulness and for the authority's power to issue it.
- We do not respond directly to requests from authorities of other countries: they can only go through official mutual legal assistance with the Netherlands, where Dutch authorities and courts review the request.
- We challenge requests we consider unlawful, overly broad or unfounded by every legal means.
- Under a binding order we disclose the minimum — only what the order expressly names and what we actually have.
- We notify you of a request before disclosing data, unless the law or the order prohibits it; if it does, we notify you once the prohibition ends.
- We have not built any “back doors”, have not given anyone our encryption keys, and have not installed software or equipment of the authorities. We will challenge any attempt to compel us to do so.
- Answers are prepared by OpenAI models (USA), so text sent to them is also subject to US law; OpenAI discloses data only on a valid legal demand.
- Every six months we publish here the number of requests received, challenged and complied with. As of 26 October 2026, requests received: 0.
Your data on the server is encrypted, but the keys are held by us. So under a binding order we are technically able to disclose data.
12. Your rights
You have the right to know what data we hold about you, to correct it, delete it, receive a copy and withdraw consent. If you believe we are violating your rights, write to privacy@rrrfirstchat.com. You also have the right to complain to the data protection authority of your country.
13. Children
Uxen is not intended for anyone under 18. We do not knowingly collect data from children. If you believe a child has created an account, write to privacy@rrrfirstchat.com and we will delete it.
14. Changes to this policy
When this policy changes, we publish the new version with a new date. We announce significant changes — new kinds of data, recipients or purposes, or longer retention — by email to your verified address (and in the bot, if linked) at least 14 days before they take effect. Clarifications that do not widen processing take effect on the day they are published.
New purposes only with your separate consent. If we want to use your data for a purpose not listed in section 4, then before that processing begins we will:
- tell you what exactly will change and what Uxen will be able to do;
- ask for your separate, explicit consent to that purpose;
- until you give it, process your data only as this policy describes.
Continued use of the service is not treated as consent to a new purpose, and such features are not switched on by default.
15. Languages
This policy is written in English, and the English text is the authoritative one. A Russian version is published at rrrfirstchat.com/privacy/ru and carries the same meaning.
If the two versions differ, the English text prevails — except where the law of your country requires otherwise. For consumers in the Russian Federation the Russian text prevails, as required by Article 8(2) of the Russian Law on Protection of Consumer Rights.
Last updated: 26 October 2026