Privacy Policy

1. Information we collect

Account data

When you create an account we collect your name, email address, authentication credentials, and (for paid plans) billing details processed by our payment provider. We also record when you accepted our legal terms and which version you accepted.

Site content

We collect and store the content you create or upload to build your site — text, images, page structure, settings, and revision history — so we can host and serve your site.

Diagnostics

When the Service encounters an error, we collect technical diagnostic data (the error, the page or action involved, browser and device information, and IP address) through our error-monitoring provider so we can fix problems.

Live-call recordings, transcripts, and screen/DOM captures

The live AI-designer call is a core feature, and it is recorded automatically — you cannot turn recording off. During a call we collect your microphone audio (recorded and transcribed), the resulting transcript and call events, and the edits produced from your requests. On website-editing calls we also collect a continuous document/DOM capture of the page you are working on plus any screen or window you choose to share (so the designer can see and edit alongside you); the initial intake call — the voice conversation before your site exists — is voice-only, with no page or screen capture. Recording (and, on editing calls, screen awareness) is a mandatory, core function and cannot be turned off. This is described in Section 3.

Usage, device, and cookie data

We and our analytics providers may collect usage and device information (such as pages viewed, actions taken, browser and device type, and IP address) using cookies and similar technologies. See Section 4.

Information from public sources (business outreach)

If we reach out to your business about the Service, the contact details we use come from information your business has published — for example, an email address on your own website or in a public business listing or directory API. We record what we contacted you about, and every such message includes a one-click opt-out; opting out is honored permanently. We do not buy contact lists and we do not use addresses collected by circumventing a website’s access controls.

2. How we use information

• to provide, operate, host, and secure the Service;
• to power AI features — transcribing your speech, understanding your requests, and generating and applying edits to your site;
• to process payments and manage subscriptions;
• to communicate with you about your account and the Service;
• to understand usage, improve the Service, and prevent fraud and abuse;
• to comply with legal obligations.

3. Live calls, screen capture, and AI processing (detailed disclosure)

The live call captures more than most web features, so this section spells it out. Recording is a mandatory, core function of every call, and screen awareness is a mandatory, core function of every website-editing call; neither can be turned off. When you are on a call:

• Your microphone audio is captured and streamed in real time, recorded, and transcribed.
• On website-editing calls, your screen/page context is captured continuously via a document/DOM capture of the editor (screen awareness), plus any screen or window you explicitly share, so the AI can see what you are working on. The initial intake call — the voice conversation that gathers your brief before your site exists — is voice-only: it is recorded and transcribed like any other call, but there is no page to see and no screen capture on it.
• This data is processed by third-party AI providers (including OpenAI Realtime) to transcribe your speech and generate responses, and by AI models (including Anthropic) to observe the page and plan and apply edits.
• We retain the audio recording, transcript, call events, and derived edits to operate the Service, support you, and improve quality. See Section 6 (Retention). If recording fails for a technical reason, the call may continue without a stored recording.

When you start a call, the Service shows a notice that the call is recorded and may be reviewed by people, with a link to this Policy; starting the call constitutes your consent to the collection and processing described in this section. For website-editing calls, the call record is also stamped with the version of our legal terms in force at the time of the call. We do not sell your call data. Our AI providers process it under their commercial API terms, which do not permit them to use it to train their models by default, and we have not opted in to any such training use.

Third parties on your call. If another person can be heard or seen, or their information appears on your screen during a call, you are responsible for having any consent the law requires (see the Terms), including in U.S. all-party-consent recording states.

4. Cookies and analytics

We use strictly necessary cookies (for example, to keep you signed in); these are required for the Service to work. We also use analytics (PostHog) to understand how the Service is used. PostHog is consent-gated: it sets no analytics cookies and PostHog’s browser SDK is not initialized until you accept via the consent banner, it stays off if you decline, and you can change your choice at any time. We use Google Tag Manager to load Google Analytics and advertising tags (Google Ads and, in future, other advertising platforms). Before you accept, Google tags run without cookies and send cookieless pings that may include the page address, referrer and browser type but no cookies or identifiers; after you accept they may set analytics and advertising cookies. When you accept, PostHog includes session replay — a recording of how you interact with pages of our app (clicks, navigation, and page content, with input values masked) — used to diagnose problems and improve the product. Separately, our servers record product events (for example, that an account signed up or a referral link was used) and send them to our analytics provider; that server-side collection uses no cookies and is not controlled by the banner. We also use Sentry for error monitoring, which receives technical data when something goes wrong. The specific cookies and technologies are listed in our Cookie Policy.

5. How we share information — subprocessors

We share information with the service providers and subprocessors below who process it on our behalf to run the Service. We require them to protect the information and to use it only to provide their services to us.

• OpenAI — Real-time voice AI (OpenAI Realtime) that powers the live call — transcribing your speech and generating the designer's spoken responses — plus AI text generation used in editing.
• Anthropic — AI models used to analyze, plan, and apply edits to your site.
• Supabase — Database, authentication, and file storage for your account and site data.
• LiveKit — Real-time audio/video transport that carries your live-call media.
• Stripe — Payment processing for paid subscriptions (card details are handled by Stripe, not stored by us).
• Vercel — Application hosting, serving, and edge infrastructure. We also route some AI requests (for example, the prompts and site context used to generate your site) to the AI providers above through Vercel's AI Gateway.
• Inngest — Durable background-job processing — for example, building your site from your intake brief and running publish jobs. Job events carry the data those jobs need, such as your brief, site content, and account/site identifiers.
• Resend — Sending transactional email such as sign-in, account, and billing notices.
• Cloudflare — Bot protection (Cloudflare Turnstile) on public forms such as signup, contact, and the waitlist. Turnstile processes browser and device signals to distinguish people from bots.
• PostHog — Product analytics to understand how the Service is used. PostHog's browser analytics (cookies/SDK) is consent-gated and not initialized until you accept (see the Cookie Policy); when you accept, it includes session replay — a recording of how you interact with pages of our app (with input values masked). Separately, our servers send product events (for example, that a signup or a referral click happened) to PostHog; that server-side collection uses no cookies.
• Google — Website analytics and advertising measurement on our marketing pages. We use Google Tag Manager to load Google Analytics and advertising tags (Google Ads and, in future, other advertising platforms). Before you accept, Google tags run without cookies and send cookieless pings that may include the page address, referrer and browser type but no cookies or identifiers; after you accept they may set analytics and advertising cookies.
• Sentry — Error monitoring and performance diagnostics. When something breaks, Sentry receives technical event data such as the error, browser/device information, and IP address so we can find and fix the problem.
• Slack — Internal operational alerts to our own staff channels. When something needs attention (for example, a billing problem or a purchased custom domain left behind after an organization is deleted), the alert message can include limited account details such as your custom domain name, organization name and identifier, and subscription identifiers — not call recordings or site content.
• GitHub — Our internal engineering issue tracker (a private repository). When our automated call-quality review detects a problem in a live call, or you submit feedback, the report it files can include your name and email address, organization and site identifiers and names, short excerpts of what was said on the call (transcript quotes), your feedback text, technical context about the edit that was being made (for example the requested change, which can quote short pieces of your site's text, and which part of the site it targeted), and call and account identifiers — not audio recordings.

We may also disclose information to comply with law, enforce our terms, or protect the rights, safety, and security of the Service, our users, or the public. If we are involved in a merger or acquisition, information may be transferred subject to this Policy.

6. Data retention

We retain your account data and site content for as long as your account is active, so we can host and serve your sites. Call audio recordings and call screen replays (the document/DOM capture of your editing session) are stored with your organization’s call records to operate, support, and improve the Service. Call transcripts, call events, and the edits derived from your requests are retained while your account is active to operate, support, and improve the Service. When you delete your organization, we delete its account data and site content as described in Section 7, except for residual copies in routine backups and records we must keep for legal, accounting, or fraud-prevention purposes.

7. Your rights — deletion and export

Depending on where you live, you may have rights to access, correct, export (portability), or delete your personal information, to object to or restrict certain processing, and to withdraw consent.

How deletion and export actually work. If you are the owner of your organization, you can download an export of its data and permanently delete the organization — including its sites, site content, and call records — from the Settings → Data page, after a typed confirmation. Self-serve export and deletion are available to the organization’s owner; if you belong to an organization you do not own, ask its owner, or contact us at support@ambermark.ai and we will handle your request. When the organization you delete was your last one, we also remove your sign-in identity; that identity removal and some storage cleanup run after the main deletion and are best-effort — if anything remains, contact us and we will complete the deletion. You can also exercise any of the rights above by contacting us at that address.

EEA/UK (GDPR)

We process personal data on the legal bases of performing our contract with you, your consent (for example, non-essential analytics), and our legitimate interests in operating and securing the Service. You may lodge a complaint with your local supervisory authority.

California (CCPA/CPRA)

California residents may request access to or deletion of their personal information and may opt out of any “sale” or “sharing” of personal information. We do not sell personal information. We will not discriminate against you for exercising these rights.

To exercise any of these rights, contact us at support@ambermark.ai. We will verify your request as required by law.

8. Security

We use reasonable technical and organizational measures to protect information. No method of transmission or storage is completely secure, and we cannot guarantee absolute security.

9. International transfers

We are based in the United States, and our providers process information there and in other countries. Where the law of your jurisdiction requires safeguards for such transfers, we rely on appropriate mechanisms — including our providers’ Standard Contractual Clauses and equivalent contractual protections.

10. Children

The Service is not directed to children under 16, and we do not knowingly collect personal information from them. If you believe a child has provided us information, contact us and we will delete it.

11. Changes to this Policy

We may update this Policy from time to time. When we make material changes we will update the “Last updated” date and, where appropriate, notify you or ask you to re-accept.

12. Contact

Questions about this Policy or your data? Contact us at support@ambermark.ai.

To report abuse or content concerns, contact abuse@ambermark.ai.

Ambermark, Inc., 954 Lexington Ave #2003, New York NY 10021.