Skip to content
Yodoca
Features FAQ
RU EN
Telegram

Personal data

Privacypolicy

Edition of 11 October 2026.

This is the operator’s document on how Yodoca processes personal data and on the safeguards in place. It is published under Article 18.1 of Federal Law No. 152-FZ of 27 July 2006 on Personal Data. The Russian text is the official version.

  1. 01Scope
  2. 02Who processes the data
  3. 03Rights and duties
  4. 04Purposes and data
  5. 05Legal grounds
  6. 06How processing works
  7. 07Who may receive data
  8. 08Databases and other countries
  9. 09The yodoca.ai website
  10. 10Three different deletions
  11. 11Retention and erasure
  12. 12Safeguards
  13. 13How to make a request
  14. 14New editions

01Scope

This policy covers the Yodoca bot in Telegram and the website yodoca.ai. It is about users of the service and visitors to the site. Staff records, if the operator hires anyone, are not described here.

Yodoca is a personal assistant in a messenger. A person sends a message, a voice note, a photo, or a document. The service replies, sets reminders, builds Word, Excel, and PDF files, makes images, and, when asked, opens a page at a link. There is no separate app. Today this works in Telegram. Another messenger is outside this policy until the policy is updated before that launch.

Chat, files, and reminders do not spend credits. Image generation past the free quota is paid in Telegram Stars, on the payment page. That does not remove the operator’s duties under 152-FZ. Telegram’s own terms are separate. Telegram sets them, not this page.

Terms mean what they mean in 152-FZ. Personal data is any information that relates, directly or indirectly, to a person. Processing is any operation on that data. The operator is individual entrepreneur Oborina Anastasia Vladimirovna. Her details are in the next section.

02Who processes the data

The personal-data operator is individual entrepreneur Oborina Anastasia Vladimirovna. She is also responsible for organising the processing.

Service
Yodoca. Bot: t.me/yodoca_bot. Site: yodoca.ai. Pages a person asks to assemble open at a link on a.yodoca.ai.
Operator
Individual entrepreneur Oborina Anastasia Vladimirovna
OGRNIP
326680000031765
Tax ID
772310565128
Address
Tambov Region, Tambov, Vysheslavtseva Street, 8
Email
hello@yodoca.ai
Responsible person
Individual entrepreneur Oborina Anastasia Vladimirovna

03Rights and duties

A person whose data is processed may:

  • learn whether their data is processed and receive the information in Article 14 part 7: confirmation, legal grounds, purposes, the data and its source, retention, how processing works, who processes it for the operator, and who received it;
  • demand correction of incomplete, outdated, or inaccurate data, and blocking or erasure if the data was obtained unlawfully or is no longer needed for the stated purpose;
  • withdraw consent and demand that processing stop where the law does not allow it to continue without consent;
  • refuse to provide data. The bot then cannot carry out the request that needed it;
  • complain to Roskomnadzor or go to court.

The operator processes data lawfully and only for purposes named in advance, does not keep more than those purposes need, answers requests, does not disclose data without a legal ground, and applies the safeguards in Article 19. People who can access the data keep it confidential under Article 7.

If an unlawful or accidental disclosure violates someone’s rights, the operator notifies Roskomnadzor: within 24 hours, the incident, the likely causes, the likely harm, and the measures taken; within 72 hours, the result of the internal review. That is Article 21 part 3.1.

04Purposes and data

Each purpose below has its own people, data, operations, and retention. Article 18.1 requires that split. The operator does not take data beyond the purpose. These records are not used for advertising, and the operator does not sell them. The purposes below are the main functionality. The ground for processing is the contract, Article 6 part 1 item 5. The contract is formed by sending the bot the command /start, as the terms describe.

Replies in Telegram

Purpose
Answer a request in the bot: look something up, draft text, or build a Word, Excel, or PDF file.
People
The person who writes to the bot. In a group the bot sees every message in the group, so messages of other participants in that chat may be processed.
Data
Telegram user id, first and last name, username if the profile has them, interface language, chat id, the message text, and files sent with the request.
Operations
Collection, recording, organisation, storage, updating, retrieval, use, deletion, and destruction. Disclosure to others only as described in the recipients section.
Retention
Indefinitely, until the user deletes it. Deleting the chat in Telegram does not erase the operator’s copy. Deletion works as described in “Three different deletions”.

Memory and reminders

Purpose
Remember what the person asks to remember, and come back to a date or an errand.
People
The bot user.
Data
Facts, names, dates, and the wording of the errand from the chat — only what the reminder or the follow-up needs.
Operations
The same operations as for a reply in Telegram.
Retention
Indefinitely, until the user deletes it. A fact is kept until it is forgotten or until a full reset. Forgetting one fact does not delete the conversation: search over old messages can still find the source. “Forget everything” asks for a code in a new message after a 15-minute preview, erases facts, the summary, and memory chunks, and hides history before that boundary from the model. Files and pages stay.

Photos and voice

Purpose
Understand a photo or a voice note and carry out the request, including making a new image.
People
The person who sent it. If other people happen to be in the frame, them too, but only within that frame and that request.
Data
The image, the voice note, and any caption. Voice and photos are used to carry out the request, not to recognise the person by face or voice. The voice recording is deleted after transcription. The text stays in the conversation.
Operations
Collection, recording, use for the reply, storage within the retention period, deletion, and destruction.
Retention
Indefinitely, until the user deletes it. A finished image and an uploaded file are kept until the person deletes them or does a full reset. Deleting a file asks for a code in the next message and lasts 15 minutes. The chat message stays. A copy already built into a published page, and a message Telegram has already delivered, are not removed by that deletion.

This is not biometric personal data under Article 11: identifying someone by face or voice is not a purpose. If a face or voice login is added later, change this policy first and get separate written consent. Do not ask people to send photos of others. Publishing a person’s image to an indefinite audience happens only when they themselves ask for a page at a link. Publication of an image is also covered by Article 152.1 of the Civil Code.

A page at a link

Purpose
Assemble a page the person asks to open at a link on a.yodoca.ai.
People
The user who asked for the page. Anyone with the link can see that page, not the bot conversation.
Data
The text and materials the person asks to put on the page.
Operations
The operations for a Telegram reply, plus access via the link. That discloses the page to whoever has the link.
Retention
Indefinitely, until the user deletes it. The link opens until the page is deleted separately in the bot or the Mini App, or until the private context is fully reset. Deleting the source file does not close the page. The page is not meant for search: the gateway sends noindex. Trash is not deletion.

Credit payments

Purpose
Take payment for a credit pack, add the credits, and handle a payment complaint.
People
The user who paid a Telegram Stars invoice.
Data
The Telegram identifier, the pack, the Stars price at the time of the order, the Telegram payment identifier, and records of credit debits and returns.
Operations
Recording, storage, use for crediting and refunds, and destruction when the law no longer requires the record.
Retention
Indefinitely. A full reset of the conversation does not delete these records.
Ground
The contract: the offer on the terms page, accepted by the command /start. Paid images are described on the payment page. Consent is not required for payment.

Personal-data requests

Purpose
Handle a request from a person or their representative and keep proof of the reply.
People
The requester and their representative.
Data
Name, contact channel, the text of the request, and details that show it is about this user: a Telegram id or username, and roughly when you talked.
Operations
Recording, storage, use for the reply, and destruction when the period ends.
Retention
Indefinitely, until the requester asks for the request to be deleted.

05Legal grounds

The grounds are Article 6 part 1 of 152-FZ. The law itself is not a ground for processing. It sets the rules for the operator.

  • Article 6 part 1 item 5: a contract to which the user is a party. The contract is the public offer on the terms page. Acceptance is sending the bot the command /start. On that ground the operator processes data for the main functionality: a Telegram reply, files, photos and voice notes, memory and reminders, a page at a link, and credit payments.
  • Item 1: consent. It is not used for the main functionality. Consent would be needed only for a purpose the contract does not cover. It would be collected separately from the terms, before that processing starts. Special categories and advertising are outside the main functionality. The boundary is on the consent page.

Where consent exists, it can be withdrawn. Silence, a pre-ticked box in someone else’s document, and merely continuing to use the bot after a new edition are not consent. Withdrawal does not stop processing that still has another ground, including the contract for the main functionality. If a new purpose needs consent, the operator asks for it separately and does not widen the purpose before that.

06How processing works

Processing is automated: the service software handles messages. A Telegram bot is not a secret chat. The operator receives the text, photos, voice notes, and files a person sends to the bot in a form that is enough to reply.

In a group the bot sees every message in the group. It processes them in order to reply in that chat. The operator does not build a separate database of participants and does not use this data for advertising. Whoever adds the bot does so under the chat’s own rules.

Special categories under Article 10 — health, beliefs, intimate life, and the others listed there — are not requested. Do not send them. There is no separate store for them, and Article 10 does not allow one without written consent. If such information does reach the conversation, it stays with that conversation indefinitely, until the user deletes it.

The service is for people over 18. Age is not checked. If the operator learns that a user is under 18, processing of that user’s data stops and the data is deleted.

The service does not make decisions that produce legal effects for a person by automated processing alone. A bot reply carries out a request. It is not a decision about credit, a job, or a benefit. That is the line in Article 16.

If a request contains someone else’s data, the sender confirms they have a right to pass it on. The operator uses it only to carry out that request.

07Who may receive data

The operator does not sell personal data and does not pass it to advertising networks. Recipients are limited to these:

  • Telegram delivers messages to the bot and back. That is the person’s choice of messenger, not processing entrusted by the operator under Article 6 part 3. The operator does not set Telegram’s rules.
  • Persons who process data on the operator’s instructions. Files and pages sit in Railway object storage. Records sit in PostgreSQL. For a reply, the text, the file fragments that are needed, an image, or a prompt go to the provider assigned to that task. The program has direct adapters for OpenAI, DeepSeek, and xAI. Speech recognition, speech synthesis, and embeddings use the OpenAI adapter when that task is assigned to it. Search and maps are separate configurable services. Each needs an Article 6 part 3 contract.
  • A public authority, where a statute requires the disclosure.
  • A person with a link to a page on a.yodoca.ai sees that page. The link does not open the bot conversation.

08Databases and other countries

When collecting personal data of people in Russia, the operator must record, organise, accumulate, store, update, and retrieve it using databases located in the Russian Federation. That is Article 18 part 5.

A user’s files and published pages are stored in Railway object storage. Conversations, memory, the profile, and payments sit in PostgreSQL.

Sending data to a foreign country, to a foreign person, is a cross-border transfer. Before it happens, the operator notifies Roskomnadzor under Article 12. The assessment depends on whether that country provides adequate protection.

The model providers in the recipients section are outside Russia. What leaves is what the reply needs: text, the file fragments that are needed, an image, a voice recording for the time it takes to transcribe it, or a prompt. That is a cross-border transfer. The operator does not own Telegram’s infrastructure.

09The yodoca.ai website

The site explains the service and leads to the bot. It has no forms, payment, or sign-up. It does not ask for a name, phone number, or email. It sets no analytics or advertising cookies of its own, the pages include no third-party counters, and the site writes nothing to localStorage. There is no separate cookie policy.

10Three different deletions

These are different operations. One does not do the work of another.

  • Memory. A single fact is deleted on its own, in the bot or the Mini App. The conversation stays. “Forget everything” hides history before a boundary from the model and erases facts, the summary, and memory chunks. Files and pages stay.
  • A published page. It is deleted on its own, in the bot or the Mini App. The link stops opening when the page pointer is removed. Deleting the source file does not close the page.
  • A full reset. It is done in the Mini App: a fresh Telegram session, the exact word RESET, and a one-time confirmation within 5 minutes. It erases the private profile, reply language, time zone, and the private context: the conversation, memory, files, and pages of that context. Group chats are not affected. The credit balance, the payment record, and the quota counters stay. Copies already delivered to Telegram, copies held by the model provider, and backups are not erased by this reset.

11Retention and erasure

A user’s data is kept indefinitely, until that user deletes it. How memory, a page, and a private context are deleted is described above. The periods below are the statutory periods when a person or Roskomnadzor demands correction, a stop, or destruction. Destruction means the data cannot be restored.

  • Inaccurate data is corrected within 7 business days from the day the person or Roskomnadzor provided the right details. During the check the data is blocked, unless blocking would infringe that person’s rights or someone else’s. Article 21 part 2.
  • Unlawful processing stops within 3 business days of discovery. If it cannot be made lawful, the data is destroyed within 10 business days and the requester is told. Article 21 part 3.
  • After the purpose is achieved, data is destroyed within 30 days, unless a contract or a statute requires longer storage. Article 21 part 4.
  • After consent is withdrawn, the same 30 days, if no other lawful ground remains. Article 21 part 5.
  • A demand to stop processing is carried out within 10 business days. The period may be extended by no more than 5 business days if the person is sent a reasoned notice. The exception is where the law allows processing without consent. Article 21 part 5.1.
  • If those periods are technically impossible, the data is blocked and destroyed no later than 6 months. Article 21 part 6.

12Safeguards

The operator applies legal, organisational, and technical measures under Article 19, matched to the risk. Conversations are not published. Deletion periods are set above.

13How to make a request

Write to hello@yodoca.ai, to the Yodoca bot, or by letter to the operator’s address in the section above. Do not send a photo of a passport to the bot.

To show the request is yours, details of how you use the service are enough: a Telegram id or username, and what you want — to know, correct, delete, withdraw consent, or stop processing. The reply uses the same channel as the request, unless you asked otherwise.

Information under Article 14 part 7 is provided within 10 business days. The period may be extended by no more than 5 business days, with a notice of the reasons. A formal request under Article 14 part 3 states the number of an identity document, the date and authority of issue, details showing you use the service, and a signature. That request may be an electronic document signed with an electronic signature. The same information may be requested again no sooner than 30 days later, if the first reply was complete.

Consent may be withdrawn in the same way it was given, or by a message to hello@yodoca.ai: any channel that shows who is withdrawing. Stopping and deletion follow the previous section. If another lawful ground remains for some data, the operator says which data and why it is still kept.

A complaint about the operator may be filed with Roskomnadzor or in court. That does not stop you from writing to the operator first.

14New editions

A new edition is published on this page and replaces the previous one from the date at the top. If a change needs new consent, the operator asks for it separately and does not keep the affected processing on the old consent while waiting. Continuing to message the bot does not, by itself, accept the new edition.

Yodoca
Investors Privacy Terms Payment Consent