Legal

Privacy Policy

1. Who we are

Tracebrain ("Tracebrain", "we") is operated by Mykola Breslavskyi, Poland. We are responsible for the personal data described on this page — the "controller", in the language of the GDPR. For anything about your data, write to [email protected].

2. What we collect

Your account

Your name, email address and profile picture, and the identifiers that sign you in. They come from the sign-up form, or from Google if you sign in with it. Passwords are handled by our sign-in provider, Clerk; we never see them.

If you subscribe

Payment is taken by Paddle, our reseller (section 7). From Paddle we receive what the purchase was — the plan, the amount, the date, and Paddle's identifiers for you, the transaction and the subscription — and record them in our server logs. We never receive your card details.

The channels you connect

Your conversations

From each connected channel we read your one-to-one chats with people: the text of the messages sent and received, when each was sent, and who the other person is — the name, username or phone number the channel shows you. Reading starts at the date you choose during setup and continues with new messages as they arrive.

We do not read group chats, channels or bots. We do not download photos, voice messages, videos or files; a message that is only media is recorded as a placeholder.

What Tracebrain writes from them

A profile for each person (who they are, what you promised each other, the birthdays and contact details that came up, and the identity document numbers - a tax number, a PESEL, a passport number - that the person gave about themselves), a note about you built the same way from your own messages, a note for each day, a task list and a list of projects. These are stored next to the conversations, in your vault.

Conversations contain sensitive things: health, money, a password somebody once sent you. They stay inside the conversation history. Card and bank account numbers are hidden before the AI reads a message, and the AI is instructed not to lift passwords, card numbers or one-time codes into profiles. An identity document number is kept only when the person wrote it about themselves and named it, and never appears in a daily note.

Technical data

Our server logs what ran and what failed. The logs of AI steps keep the model's answers, which can quote your conversations; they stay on the same server as your vault and are deleted with it. Our hosting and sign-in providers keep their own request logs: IP address, browser, time, and the address of the page requested.

How you use the website and the app

Both use PostHog, a product analytics service, to show us which pages are opened, which buttons are pressed and what breaks. It receives the address of each page, your browser and device, your IP address and the rough location it gives, and a random identifier kept in your browser. Once you are signed in to the app, it also receives your account identifier — a random code, not your name or email address — and counts actions such as starting an import or adding a task.

In the app, the name in the address of a person's or a project's page is removed from every event before it is sent, the text of what you click is not recorded, and an error is recorded by its place in the code without its message.

Sessions are recorded too — how a page is scrolled and clicked. Everything in a recording that came out of your vault is masked: a name, a line of a note, a promise, a quote from a message. The app's own words, a heading or a button, stay readable, because they are ours and not yours. Nothing you type is kept, no request the app makes is kept, and a pairing QR code is not recorded at all.

One thing a recording does keep is the address of each page, and the address of a person's page contains the name of their note. Apart from that address, PostHog receives none of your conversations, none of your notes, and none of the names of the people in them.

This website — every page outside the app — may be recorded in the same way, but nothing here comes from a vault, so what is hidden here is only what you type into a form.

This website, and not the app, also uses Google Analytics, but only if you accept Google's cookies in the cookie banner. Google then receives the pages you open, your browser and device, the rough location your IP address gives, and a random identifier kept in a cookie, so that we can see which pages bring people here. Until you accept, Google receives only an anonymous signal that a page was opened, with no cookie and no identifier.

When you write to us

Your email and whatever you put in it. The "request the deck" form on the investor page sends the address and message you enter to our inbox through FormSubmit, a form-forwarding service. The early-access form on the company memory page does the same with your address and the answers you pick (the size of your team, the messenger it works in and your role); our analytics record those answers, never the address.

3. Signing in with Google

If you choose "Continue with Google", Google tells us — through Clerk — your name, email address, profile picture and a Google account identifier. We ask for the basic sign-in scopes only (openid, email, profile). We never ask Google for access to your Gmail, Drive, Calendar or Contacts.

Tracebrain's use of information received from Google APIs will adhere to the Google API Services User Data Policy, including the Limited Use requirements.

4. Other people in your conversations

Your conversations hold the words and details of the people you talk to, and they have not signed up for Tracebrain. We process their information for one purpose: to build your memory, for you. We never contact them, never build profiles of them for anyone else, and never combine what one Tracebrain user knows about a person with what another does.

If you appear in someone's Tracebrain vault and want to exercise your rights, write to [email protected] and we will take it up with the account holder.

5. How we use it

What we never do:

6. Legal bases

Where the GDPR applies, we rely on:

7. Who processes it for us

These companies process data on our behalf, under contract, and only to provide the service: your account, your vault and the conversations in it, and, for PostHog and Google, the record of how the website and the app are used. Paddle is listed here too, though it sells the subscription in its own name as merchant of record rather than on our behalf: it decides for itself how it handles what it needs to take your payment. This is the whole list for that data. Requesting the investor deck or early access is separate and is not part of the product: either form sends only what you enter in it to FormSubmit, the form-forwarding service named in section 2.

Hetzner Online GmbHGermany

Runs the server in Germany that holds every vault and the programs that read your channels. Your conversations and everything written from them are stored here. Privacy policy · ISO/IEC 27001:2022

Supabase, Inc.Frankfurt, Germany

Hosts the database that records which account owns which vault and which channels are connected, including the sessions that let us read them. It holds no messages. Privacy policy · SOC 2 Type 2 · ISO 27001

Clerk, Inc.United States

Handles sign-up, sign-in and sessions: your name, email address, profile picture and, if you use one, your password. It never sees your conversations. Privacy policy · SOC 2 Type 2

Paddle.com Market LimitedUnited Kingdom

Sells the subscription as our reseller and merchant of record: it takes the payment, charges any tax that applies and sends the receipt. It receives your email address, billing country and postcode, your payment details, and an identifier that ties the purchase to your Tracebrain account. We never see your card, and it never sees your conversations. Privacy policy

Anthropic, PBCUnited States

Provides the Claude models that write profiles, daily notes and tasks. To do that it receives the text of the conversations being summarized, and returns the result. Privacy policy

Vercel Inc.United States

Serves this website and the web app. What you open in the app passes through its servers on the way to your browser and is not stored there; it keeps request logs such as IP address, browser and the address of the page. Privacy policy · SOC 2 Type 2 · ISO 27001:2022

PostHog Inc.Frankfurt, Germany

Product analytics for this website and the app: the pages you open, the buttons you press, your browser, device, IP address and the rough location it gives, and, once you are signed in, your account identifier, never your name or email. In the app it also counts what you do, such as starting an import or adding a task, records errors by their place in the code, and records how a page is used - with everything that came out of your vault masked in the recording, except the address of the page, which contains the name of a person's note. Apart from that address it receives none of your conversations, none of your notes and none of the names of the people in them. Privacy policy · SOC 2 Type 2

Google Ireland LimitedIreland, with transfers to the United States

Google Analytics, on this website only and only if you accept Google's cookies in the cookie banner: the pages you open, your browser and device, the rough location your IP address gives, and a random identifier kept in a cookie, so that we can tell which pages bring people here. It never runs in the app, and it never receives your conversations, your notes or anything you type. Privacy policy

Anthropic receives your conversations only to return the result to us, and handles what it receives under its own terms and privacy policy, linked above. Each AI run is confined to your vault alone and has no access to anyone else's.

Telegram and WhatsApp are not our processors, and neither is Google when you sign in with it. They are the services you already use, and your use of them stays under their own policies.

If you turn on notifications, they reach your device through the push service your browser maker runs (Apple, Google or Mozilla). A notification carries the date of a note that is ready, never a name or a message. Turning notifications off in the app, or deleting your data, removes the address we keep for your device.

We disclose data beyond this list in two cases only: when the law compels us — and we will tell you, unless we are forbidden to — and if Tracebrain is ever acquired, in which case your data moves under the commitments on this page and you hear about it first.

8. Where it is kept, and how it is protected

Your vault is on a server in Germany. The registry of accounts and channel sessions is in Frankfurt, and so are the analytics PostHog keeps. Clerk, Anthropic, Vercel and Google process data in the United States and elsewhere, and PostHog is a U.S. company; those transfers rely on the safeguards the providers offer: the European Commission's Standard Contractual Clauses, or certification under the EU–U.S. Data Privacy Framework.

Everything is encrypted in transit: between your browser and the app, between the app and our server, and between our server and the database, which also verifies the database's certificate. Each person has a separate vault under a random name, and every request is verified against your signed-in session — its signature, issuer, expiry and origin — before anything is read. The database accepts connections from our server only, and its tables are closed to every other role.

An AI run works inside your vault alone, has no shell, and what it proposes for your tasks and projects is verified by ordinary code before it is saved. The registry that holds your Telegram session is encrypted at rest with AES-256, and sessions and keys are never written to logs.

The certificates named in section 7 are our providers' own, published by them: the data centre that holds your vault is certified to ISO/IEC 27001, and our sign-in, database, hosting and analytics providers are audited to SOC 2 Type 2. Tracebrain itself holds no certification yet.

No system is perfectly secure. If a breach affects your data, we will tell you and the authorities without undue delay, as the law requires.

9. How long we keep it

10. Your rights and choices

You can ask us for a copy of your data, to correct it, to delete it, to restrict or object to how we use it, and to hand it to you in a portable form — your vault is plain Markdown files, and we send it as an archive. Write to [email protected]; we answer within one month. You can also delete your vault and your channel sessions yourself, at any time, with Delete my data (section 9).

If you think we have handled your data wrongly, you can complain to your data protection authority. In the EU, the EDPB lists them. We would rather hear from you first.

11. Cookies

The app uses the cookies that keep you signed in, set by Clerk. This website and the app also use PostHog, which keeps a random identifier in a cookie and in your browser's local storage, so that it can tell a returning browser from a new one. The app uses no advertising cookies. This website sets Google's analytics cookies only if you accept them in the banner; it remembers your answer in your browser, and Cookie settings at the foot of every page lets you change it at any time. Rejecting them, or withdrawing later, removes the ones already set. The app also keeps its own code and an offline page in your browser, so it opens quickly and says when you are offline; it never stores your notes or messages there.

12. Age

Tracebrain is for adults. You must be at least 18 to use it, and we do not knowingly collect data from anyone younger.

13. Changes to this policy

When this page changes, the date at the top changes with it. If a change affects how we use your data, we will tell you by email before it takes effect and, where the law requires, ask for your consent.

14. Contact

Mykola Breslavskyi, Poland · VAT PL8993038838 · [email protected]. The Terms of Service are the other half of this agreement.