AskAIs is now telmini. Your account, data and plan stay the same. Learn more

telmini
Get started

Privacy Policy

Last updated: October 30, 2026

This Privacy Policy explains how telmini (formerly AskAIs) collects, uses, shares, and protects personal information. The telmini service is operated by MiniCode LLC, located at 30 N Gould St Ste R, Sheridan, WY 82801, USA (“telmini”, “we”, “us”), the data controller for personal information about our account holders and website visitors. It applies to our websites at telmini.com, askais.com, and cs.askais.com and to the telmini service.

Public website measurements

When enabled, our first-party website counter records daily totals for page views and clicks on sign-up, sign-in and App Store links. These totals are grouped by broad website section and language, not by visitor or account. The counter does not store full URLs, query strings, message content or raw IP addresses, and does not set tracking cookies. It respects browser Do Not Track and Global Privacy Control signals. A keyed, short-lived hash of a network address is used for rate limiting and expires within two minutes. Aggregates older than 90 days are pruned during collection; remaining aggregates may be retained while collection is disabled. Service authentication and operational security logs are separate from this counter.

1. Controller and processor roles

We act as a data controller for personal information about our customers (account holders) and our own website visitors. When our customers use telmini to communicate with their end users, we process those end users’ personal information as a data processor, on our customers’ behalf and under their instructions; in that case the customer is the controller. This policy describes our practices as a controller. Our processing on behalf of customers is governed by our agreement with them, including a Data Processing Addendum available on request. We keep each customer’s data logically separated from that of other customers (multi-tenant isolation). The Service may also be white-labeled, so an end user may see only the customer’s brand and may not be aware that telmini provides the underlying technology.

2. Information we collect

  • Account information: your name, profile photo, email address, password, workspace and company details, and role.
  • Billing information: your plan, transaction history, and billing identifiers. Card details are collected and stored by Stripe, not by us.
  • Usage and device data: log data, IP address, browser and device type, pages viewed, and actions taken in the Service.
  • Connected AI client data: when you authorize ChatGPT, Codex, Claude, or another MCP-compatible client, we process the client name, your telmini user and workspace identifiers, granted permissions, connection and revocation timestamps, and limited MCP call metadata such as tool name, outcome, duration, and client platform. We do not store MCP tool arguments, prompts, tool results, passwords, full API keys, or plaintext OAuth tokens in the MCP usage log.
  • Support communications: messages and information you send when you contact us.
  • Customer Data: conversations, contacts, knowledge-base documents, and other content that our customers and their end users submit, which we process as a processor (see “Controller and processor roles”). Depending on how a customer configures the Service, this may also include an end user’s IP address and the approximate location derived from it (such as country and city), the platform and device they use to reach the customer, any custom attributes the customer records about them, membership or account fields the customer chooses to store, voice messages and their transcripts, and translations of messages that agents choose to translate.
  • Team chat and Notebook: messages, voice messages, and attachments that members of a workspace send each other in team chat (the workspace’s team channel, direct messages between two members, and groups that members create, with each group’s name and photo), and the notes and checklists they write in Notebook, either personal (visible only to their author) or shared with the team. This is Customer Data of that workspace. Team chat messages and notes are stored encrypted on our servers with a key unique to each workspace; this is server-side encryption, not end-to-end encryption. Notes can include images that agents add in the app. Before uploading an image, the app removes its location and other photo metadata. Images are stored encrypted on our servers like the notes themselves, and are deleted when the note is permanently deleted. If an agent turns on Telegram or email reminders for Notebook to-dos, reminders for to-dos assigned to them in team notes that are due soon or due now are sent to them through Telegram or our email delivery provider, and contain the note title, the to-do text, the due time, and a link to the note. To send Telegram reminders, we keep the Telegram account they connected (its chat ID and username) until they disconnect it.
  • Cookies and similar technologies (see “Cookies and tracking”).

3. How we use information

  • to provide, operate, and maintain the Service;
  • to authenticate users and secure accounts;
  • to establish, operate, audit, and let you revoke authorized MCP connections;
  • to process payments and prevent fraud;
  • to provide AI-assisted replies, translation, voice dictation, and related features (see “Third-party AI services”);
  • to respond to support requests;
  • to improve the Service;
  • to send service-related and, where permitted, marketing communications;
  • to comply with legal obligations and enforce our Terms.

4. Legal bases (EEA and UK)

Where the GDPR or UK GDPR applies, we rely on the following legal bases: performance of a contract (to provide the Service and process billing); our legitimate interests (to secure and improve the Service); your consent (for example, certain cookies or marketing); and compliance with legal obligations. You may withdraw consent at any time where processing is based on consent.

5. Payments

We use Stripe to process payments. Stripe processes payment data under its own terms and privacy policy, which you can read at https://stripe.com/privacy.

6. Connected AI clients and MCP

If you choose to connect an MCP-compatible AI client, that client can access only the telmini workspace and permission scopes shown on the authorization screen. Depending on the permissions you grant and the tool you ask the client to use, relevant Customer Data may be transmitted between the client and telmini to perform your request. Read and write tools are identified separately, and external or destructive actions may require confirmation in the client.

OAuth authorization codes are short-lived and single-use. Access and refresh tokens are random credentials returned only to the authorized client; telmini stores only cryptographic hashes of those tokens. You can review and revoke connections from your telmini MCP connection settings. Revocation immediately prevents further use of associated tokens.

7. Third-party AI services

Some features of telmini use a third-party AI service to process content. All of these AI features — on our website and web dashboard and in our apps — use OpenAI only, which acts as our subprocessor; we do not send this content to any other AI provider. We send OpenAI only the content a feature needs, and only when that feature is used:

  • AI replies. When a workspace turns on automated replies that use our platform-hosted AI, we send OpenAI the visitor’s messages and the recent conversation, images the visitor sent in it, relevant excerpts from the workspace’s knowledge base, and contact details the workspace already holds that help answer the question, such as the visitor’s name, device, and approximate location, so that it can draft a reply. The visitor’s IP address is not included. The replies are stored in the conversation like any other message.
  • Translation. When an agent chooses to translate a message or a reply draft, in the web dashboard or in the mobile app, we send OpenAI the text of that message or draft, the language to translate it into, and matching terms from the workspace’s glossary. The translation of a received message is saved with that message so that it does not have to be translated again; reply drafts and their translations are not stored on our servers.
  • Voice dictation. When an agent dictates a reply, in the web dashboard or in the mobile app, our servers forward the recording to OpenAI to convert it into text. We do not store the recording; the text is placed in the reply box for the agent to review before sending.
  • Voice message transcripts. When a workspace has turned on voice-to-text (speech-to-text), voice messages and audio files sent in its conversations and team chat, including those recorded in our apps, are sent to OpenAI to produce a written transcript, which is stored with the message. The voice message itself is delivered only to its recipients.
  • Team chat translation. When an agent translates a teammate’s message in team chat, or turns on automatic translation there, we send OpenAI the text of that message (or of new messages from teammates) and the language to translate it into. The translation is saved with that message in the workspace.
  • AI Settings Assistant. When an agent chats with the assistant, in the web dashboard or in the mobile app, we send OpenAI what the agent types, the earlier messages of that assistant conversation, the agent’s role in the workspace, and the workspace information the assistant looks up to answer or to prepare a change — such as workspace settings, FAQs, keyword rules, canned replies, the list of knowledge-base documents, and conversation statistics. Nothing in the workspace is changed until the agent confirms. The assistant conversation is kept with the workspace’s data.
  • Notebook “Write with AI”. When an agent asks AI to write a note, we send OpenAI the customer conversation (including its internal notes), team chat (with teammates’ names), note or description the agent chose, and the instructions the agent types, so that it can draft the note. Nothing is saved until the agent saves the draft. Notes are not sent to any AI service unless an agent uses “Write with AI”.
  • Handing a conversation back to AI (“Resume AI handling”). When an agent hands a conversation back to AI, we send OpenAI the recent messages of that conversation (what the customer and the team wrote), images the customer sent in it, and contact details the workspace already holds — the customer’s name, a partly masked email address, device, and approximate location, as well as the app or browser they use, their account tier, and their earlier conversations with the team — so that the workspace’s AI can reply. The AI’s replies are stored in the conversation like any other message.
  • AI help in Contact Support. When an agent uses Contact Support in our app, we send OpenAI the messages, images, and files they send there, together with their name, work email address (partly masked), workspace, app and device details, approximate location (estimated from their IP address), and earlier support conversations, so that AI can suggest or write replies from our support team. We keep that support conversation as our support record.
  • AI learning and knowledge base. When a workspace has turned on AI learning and an agent sends resolved conversations to it, we send OpenAI those conversations so that it can draft question-and-answer cases, which the workspace reviews before its AI uses them. Documents, web pages, and Q&A a workspace adds to its knowledge base, including in the app, are sent to OpenAI to be indexed, so that relevant excerpts can be found for AI replies; when an agent uses the test chat, in the web dashboard or in the app, the test messages they type are also sent to OpenAI to write the replies, and nothing is sent to customers. AI setup. A workspace can set up its AI in the web dashboard or, for members allowed to, in the app (Tools > AI setup). What is sent: the instructions, greeting, procedures, and Q&A the workspace writes, which are sent with each AI reply; short-term updates the workspace submits, so that the AI can summarize them before using them in replies; and, once a week, resolved conversations of the workspace, so that the AI can draft Q&A suggestions — only the messages the customer could see, with email addresses, phone numbers, order numbers, and similar identifiers masked — and the first question customers asked in conversations the AI did not resolve, so that similar questions can be grouped into a knowledge-gap report. Suggestions are not used until someone in the workspace approves them, and the knowledge-gap report adds nothing to the knowledge base. If a workspace sets up custom AI tools, the AI may send details from the conversation, such as an order number, to the workspace’s own systems that the workspace configured, and the result is sent to OpenAI to write the reply; the credentials for those systems are stored encrypted and are never sent to the AI. Kept: with the workspace’s data until the workspace deletes it; updates stop being used when they expire. Sent to: OpenAI.

The only third-party AI service provider we currently use for these features is:

  • OpenAI: AI replies to customers (including after a conversation is handed back to AI, and AI help in Contact Support), translation (including team chat translation and automatic translation), voice dictation, voice message transcripts, the AI Settings Assistant, Notebook’s “Write with AI”, learning from past conversations, knowledge-base indexing and search, the other AI tools in the web dashboard. See OpenAI’s API data usage policy.

We update this list, and the notices in our apps, before a different provider receives any content.

Nothing is sent for these features unless they are used: AI replies, voice message transcripts, AI learning, and knowledge-base indexing only when the workspace has turned them on, and the other features only when an agent chooses to use them. AI features are part of the workspace tool: a workspace’s owner or administrators decide which AI features are available to its agents, and can turn them off for everyone. The telmini apps for iPhone, iPad, Mac, Android, and Windows follow those workspace settings and have no separate AI switches. Team chat messages, the photos, videos, and files agents send, and Notebook notes are not sent to any AI service unless an agent uses one of the features above on them, such as translating a message or “Write with AI”; voice messages are sent only for transcripts, when the workspace has turned on voice-to-text. Translated messages and drafts can include information that a workspace’s customers have shared, such as their name or order details; the workspace decides whether its agents may use translation.

OpenAI processes this content only to return the reply, translation, or transcript to us. According to OpenAI’s API data usage policy, content sent through its API is not used to train OpenAI’s models by default, and OpenAI may retain it for up to 30 days to monitor for abuse.

We engage OpenAI under written terms, including a data processing addendum, that require it to provide the same or equal protection of personal data as described in this Privacy Policy; under those terms, content we send is not used to train OpenAI’s models, and OpenAI may keep it for up to 30 days to monitor for abuse before deleting it. We and OpenAI use this content only to provide the feature that was used; we do not use it for advertising or to train AI models, and we do not sell it.

A workspace can instead connect its own AI provider account (its own API key) for AI replies. In that case the content for those replies is sent to the provider the workspace chose, under the workspace’s own agreement with that provider.

8. Crash and error reporting

We use Sentry, a service provided by Functional Software, Inc. in the United States and acting as our service provider, to receive crash and error reports from our servers, from our website and web dashboard, and from the telmini apps, so that we can diagnose errors and fix them. We use these reports only to diagnose problems, so that we can provide better service and a better user experience; they are not used for analytics, tracking, advertising, or AI. A report describes the error itself: for example, the type of error and where in our software it occurred, the software or app version, and the type of browser or device involved.

Personal data is removed before a report is sent. Reports do not include message content, email addresses, passwords, sign-in tokens or other credentials, or IP addresses, and web addresses in a report are shortened so that search terms and other parameters are not sent. Reports are not linked to your name or email address; if we choose to, a report from our servers or website may carry a one-way coded identifier that shows whether two errors affected the same account but does not reveal who you are. Reports from the app are not linked to your account at all.

Crash and error reports are generally deleted automatically within 30 days; we may delete them sooner once the problem is fixed. In the app, crash reports are on by default, and you can turn them off at any time under Me › Error Reports › Send Crash Reports Automatically. The app’s crash reports are described in more detail in the telmini Mobile App Privacy Notice (see “Mobile apps” below).

We do not use analytics tools. Third-party services we use process data under their own terms and privacy policies.

9. How we share information

We do not sell your personal information. We share it only as follows:

  • Service providers and subprocessors that help us operate the Service, such as hosting, object storage for files and attachments (such as Amazon S3 or Cloudflare R2), payments (Stripe), AI model providers (see “Third-party AI services”), messaging channel providers (such as Telegram) where a customer connects those channels, Telegram for Notebook reminders an agent turns on, email delivery (Resend), and crash and error reporting (Sentry; see “Crash and error reporting”), under contracts that require them to provide the same or equal protection of personal data as described in this Privacy Policy;
  • Legal and safety: to comply with law, respond to lawful requests, or protect the rights, property, and safety of telmini, our users, and the public;
  • Business transfers: in connection with a merger, acquisition, financing, or sale of assets;
  • With your consent or at your direction.

10. Cookies and tracking

We use cookies and similar technologies to keep you signed in, remember preferences, and secure the Service. Third-party services we use may set their own cookies under their own terms and privacy policies. You can control cookies through your browser settings, though some features may not work without them.

11. Data retention

We retain personal information for as long as your account is active and as needed to provide the Service, comply with legal obligations, resolve disputes, and enforce our agreements. Customer Data is retained according to our customers’ instructions and our agreements with them, and is deleted within a reasonable period after account termination.

OAuth authorization codes expire within minutes. OAuth access tokens normally expire within one hour and refresh authorization normally expires within 30 days unless revoked sooner. We retain connection records and limited MCP usage metadata while needed to operate, secure, and audit the connection and account. Revoked credential hashes and security records may be retained for a limited period to detect replay or abuse; they cannot be used as plaintext credentials.

When a message is deleted, we make reasonable efforts to also remove any associated files and attachments from our object storage, although residual copies may persist for a limited time in backups.

12. Security

We use technical and organizational measures to protect personal information, including encryption in transit, access controls, and reliance on PCI-DSS-compliant providers for card processing. Team chat messages and Notebook notes are also encrypted at rest with a key unique to each workspace. No method of transmission or storage is completely secure, and we cannot guarantee absolute security.

13. International transfers

We are based in the United States and may process personal information there and in other countries. Where required, we use appropriate safeguards, such as Standard Contractual Clauses, for transfers from the EEA, UK, or Switzerland.

14. Your rights

Depending on where you live, you may have rights to access, correct, delete, or port your personal information, to object to or restrict certain processing, and to withdraw consent. Residents of the EEA and UK may also lodge a complaint with their local supervisory authority.

California residents have rights under the CCPA/CPRA, including to know, delete, and correct personal information and to opt out of the sale or sharing of personal information, and, as noted above, we do not sell personal information. We will not discriminate against you for exercising your rights.

To exercise your rights, email support@telmini.com; we may need to verify your identity. If the information is held on behalf of one of our customers, we will refer your request to that customer.

15. Children’s privacy

The Service, including our mobile app, is a business tool for companies and their support staff and is not directed to children. We do not knowingly collect personal information from children under 16. If you believe a child has provided us personal information, contact us and we will delete it.

16. Data Processing Addendum

Business customers who require a Data Processing Addendum (DPA) for GDPR or UK GDPR compliance can request one at support@telmini.com.

17. Changes to this policy

We may update this Privacy Policy from time to time. We will post the updated version here with a new date and, where appropriate, provide additional notice.

18. Mobile apps

We publish the telmini agent app, a mobile app that lets a customer’s support agents work from a phone. Everything above applies to the app. Because the app also handles a few things that the web service does not — push notifications, the camera, microphone, and photo picker, device records used for remote sign-out, and in-app account deletion — those, together with the customer details (including IP address) that the app shows to agents, are described separately in our telmini Mobile App Privacy Notice, which supplements this policy rather than replacing it. That notice also explains team chat, Notebook, and how the AI features in the app work. The main points about the data the app handles are also summarized below.

Team chat and groups. In the app, members of a workspace can talk to each other in the workspace’s team channel, in direct messages between two members, and in groups. Workspace owners and administrators can create groups, and other members can too unless the workspace has turned this off for them. A group has a name, an optional photo, and members with the role of owner, admin, or member, and only its members can read it. Team chat messages are stored on our servers, encrypted with a key unique to each workspace (server-side encryption, not end-to-end encryption). Group names and photos, and notices about changes to a group, such as who added whom, are stored with the workspace’s data but are not covered by this encryption.

Privacy mode. Where a workspace has made it available, a team chat can be put in privacy mode. While it is on, notifications for that chat show the sender’s name and “New message” instead of the message text; copying, forwarding, and downloading its messages and attachments are blocked; and on iPhone, screenshots and screen recordings of the chat show a blank area. We record who turned privacy mode on or off and when, and, if the workspace has turned on screenshot notices, who took a screenshot of the chat and when, without any image or message content.

Your profile and photos. In the app you can change the name on your profile, and take or choose a profile photo for yourself or a photo for a group, using the camera (on iPhone and iPad) or the system photo picker, which gives the app only the photo you pick. The app crops the photo and converts it to JPEG on your device, and it is stored in your company’s workspace on our servers. Members of the workspace see your name and profile photo, updated in real time, and customers see them next to the replies you send them; a group’s photo is shown to that group’s members.

Inviting teammates. Workspace owners and administrators, and other members the workspace allows, can invite a colleague, in the app or in the web dashboard, by entering the colleague’s email address and a role. The email address is sent to our servers, and we send the invitation email to it on the workspace’s behalf, with the inviter’s name, the workspace’s name, and a link to join that expires after 7 days. We keep the invitation record (the email address, role, who sent it, when it expires, and whether it was accepted or revoked) with the workspace’s data, together with an entry in the workspace’s audit log.

Widgets. So that the app’s widgets can show it, the app keeps a small summary on your device, in storage shared only between the app and its widgets: unread counts, the names of recent customer conversations and team chats, to-do titles, recent note titles, and, if message previews are on, short message previews. Previews from chats in privacy mode are never included. The widgets do not connect to the internet, and the summary is not sent anywhere. It is cleared when you sign out, and you can limit what the widgets show in the app’s settings.

Notes saved on your device. On iPhone, iPad, and Mac, if a change to a note can’t be saved to our servers right away, for example while you are offline, the app keeps it on your device, encrypted with a key stored in the device’s keychain. It is deleted once the change is saved. This copy is not included in device backups, and it is also removed when you sign out or leave the workspace.

Deleting conversations. “Delete for me” hides a customer conversation or team chat from your own list only: nothing is deleted, other people still see it, and it reappears when a new message arrives. “Delete for everyone” is available only if the workspace has turned it on and you have permission to use it. It immediately and permanently deletes the conversation’s messages and attachments, including their translations and transcripts, for everyone; deleting a group this way also removes the group. We keep a record in the workspace’s audit log of who deleted the conversation and when, with counts but no content, and we keep copies of any messages that had been reported for review, as evidence of that report.

Announcements and update checks. The app asks our servers for announcements while you are signed in, sending the platform and the app’s language, and we record which announcements you have seen so that they are not shown to you again. To check for updates, the app asks our servers for the latest version, sending only the platform and the app’s language, without your sign-in token or any account or device identifier; the comparison with the version you have installed is made on your device.

Report a Problem. Under Me › Error Reports, you can send us a description of a problem, up to three screenshots you choose, the app and device details, and, if you leave that option on, recent error log lines from which the app removes message content, email addresses, tokens, and IP addresses. The report reaches our support team as a Contact Support conversation linked to your name, work email address, and workspace. We use it only to answer you and fix the problem, and it is deleted automatically 7 days after we resolve it. Automatic crash reports are described in “Crash and error reporting” above.

Android and Windows apps. The telmini agent app is also available for Android, on Google Play and as a download from our website, and for Windows, as a download from our website. Starting with version 1.3.1, they have the features summarized above, apart from widgets on Windows, and they are described in “The Android app” and “The Windows app” in the Mobile App Privacy Notice. In short: they do not use any push notification service and register no push token; while you are signed in, the app keeps an encrypted connection to our servers and shows alerts itself, and alerts about team chat messages never include the message text. Team chat in these apps uses the server-side encryption described above, not end-to-end encryption. In privacy mode, Android blocks screenshots and screen recordings of the chat, and on Windows the chat appears blank in screenshots and screen recordings. The Android widgets keep their summary only on your device. The update check tells our servers whether the Android app came from Google Play or from our website, so that it can point you to the right place to update. If you installed the Android app from Google Play, the app may ask Google Play to show its rating dialog; any rating or review is handled by Google Play under Google’s own terms, and the app does not receive it. The Android and Windows apps offer no purchases.

To delete your account, either in the app or by email, and to see which data is deleted and which is kept, see Delete your telmini account.

19. Contact

For privacy questions or to exercise your rights, email support@telmini.com or write to MiniCode LLC, 30 N Gould St Ste R, Sheridan, WY 82801, USA.