Privacy policy

Last updated: 11 October 2026

ExportMyDocs.com provides tools to import, review and export Google Docs and Microsoft Word documents. The upload page identifies the active processing mode before you choose a file. Browser mode processes local documents in the page. Server mode uploads DOCX files for processing and temporary storage. Google Docs link imports are available in both modes and also involve Google and our server.

Who is responsible

This is the controller and your contact for support and privacy requests. See the Impressum for provider information.

Files you upload

Browser mode: local DOCX and PDF files are read and processed in the open browser page. The application does not upload their contents for conversion in this mode. Document text, images, comments and other supported content remain in the page while you work. Closing the page releases this working state; downloads remain wherever you save them.

Server mode: a local DOCX is uploaded directly to private Supabase storage. Our server reads supported text, comments and embedded images to prepare the selected export. Vercel runs the processing code; Supabase stores the source, result, queue, progress and storage accounting. Files are not publicly listed. Download access is checked against your guest processing session or signed-in account and uses a temporary signed link.

Accepted source files have an initial 24-hour lifetime. Result retention is fixed when the server admits an export: a Free result is available for 24 hours from its first successful completion, and a Weekly or Lifetime result for 7 days. Older results keep their recorded expiry when your plan changes. Creating or opening a shared link does not extend that deadline. Access expires at the recorded expiry time and automatic cleanup is scheduled afterward; physical deletion may be delayed if cleanup is unavailable. Unfinished uploads also expire automatically. Processing records, including random session and job identifiers, sizes, checksums, timestamps, status and error codes, can remain after file cleanup; these records currently have no automatic deletion schedule. Provider backups and logs follow the provider's retention settings. Download your results before expiry and keep your original files.

Your browser or operating system may retain its own history, cache or downloaded files. We do not send document content to an AI model for conversion.

Text recognition in images

When you open the image gallery, text detection runs locally in your browser. Manual detection and text extraction also run locally. The browser downloads Tesseract recognition code and language models from jsDelivr. Those requests expose normal connection metadata, such as your IP address, to the CDN; they do not contain document images or recognized text. Language models may be cached in your browser for later use. Images, recognition results and your corrections remain in the open page and are not stored in that model cache. Download results before closing the page or changing the document. See jsDelivr’s privacy policy for its handling of CDN requests.

Google Docs imports and permissions

A public-link import sends the document ID and any resource key to our server. In browser mode, the server requests a DOCX or PDF export from Google and passes the bytes to your browser for that request. In server mode, it records an import task, fetches the DOCX from Google into private Supabase storage and prepares the downloadable ZIP. The same temporary file retention applies as for uploaded documents.

If you connect Google, ExportMyDocs.com requests the Google Drive read-only permission (drive.readonly). This permission is broader than access to a single selected file. The interface uses it to request documents whose links you provide: their names, exported content and available comments, author names, dates, quoted passages, replies and resolved status. In browser mode, connected imports fetch document content and comments directly from Google into your browser. In server mode, our server fetches the content and, when you select the Google comment register, the available comments into private storage. The register is included as separate JSON and CSV files in the ZIP. ExportMyDocs.com does not modify the source document.

Google authorization codes and tokens pass through our server to establish, validate or renew your connection. In browser mode, the server returns an access token to the browser for the requested import. In server mode, session responses contain connection status without an access token; background imports use the encrypted credentials described below. You can use local uploads without granting Google access. Tool pages load Google's sign-in script, so Google receives a network request even before you choose to connect.

Google data is used to provide the import, preview and export features you request. We do not sell it, use it for advertising or transfer it to train generalized AI or machine-learning models. ExportMyDocs.com's use and transfer of information received from Google APIs follows the Google API Services User Data Policy, including its Limited Use requirements.

Optional accounts and payments

Where accounts are enabled, Supabase Auth handles sign-in using Google identity or an email code. We keep an internal account identifier, its authentication-user link, account state and links to its document ownership records. Account and authentication records include your email and verification status. Authentication sessions use Secure, HttpOnly cookies in production; the application does not store authentication tokens in browser local storage. Sign-in email is delivered through Supabase Auth's configured email service. Google account sign-in does not itself grant Google Drive access.

You can opt to attach documents from this browser's guest session to an account and then confirm the target account. That attachment allows the account to access those records across devices and prevents the old guest cookie from accessing them. Payment activation is separate: an unclaimed guest purchase can be attached only after sign-in verifies the email recorded at Checkout. Matching that email does not attach the guest's documents. An existing purchase's account owner is not changed when its billing email changes.

Where payments are enabled, Stripe collects payment details through Checkout and its billing portal. Our application does not receive or store card numbers or card security codes. We store purchase, Customer, subscription, invoice and PaymentIntent identifiers; the completed Checkout email; product and policy identifiers; amounts, currencies, payment and refund states; payment-document links; and timestamps. Stripe metadata contains internal purchase and policy identifiers, not document content or document links. These records support payment activation, access limits, payment history, cancellation, reconciliation and support. They are kept separately from temporary document files and are not automatically deleted when a file expires or an account is removed.

A signed, HttpOnly browser-context cookie lets the browser check its own Checkout status without signing in; it does not grant account or document access. Browser session storage may keep an opaque purchase identifier and retry keys. If purchase email notifications are configured, Resend receives the Checkout email and a purchase notification with a link to ordinary account sign-in. That link does not sign you in. These purchase notifications are separate from sign-in codes sent through Supabase Auth.

Processing session and recovery

Guest server processing uses a signed, Secure, HttpOnly cookie to control access to uploads and jobs. Signed-in processing uses the verified account and its document ownership records. Browser storage keeps the current document filename, size, upload or import identifier, job identifier, request key and recovery stage, separated by guest or account identity. For Google imports it also keeps the document ID, optional resource key and selected import options so a lost response can be retried without creating another import; it does not store Google tokens. This lets the same browser resume checking an upload or job after a reload. Clearing cookies or site storage can remove guest access and local recovery information; documents already attached to an account remain accessible after sign-in while they are retained. Starting another document clears the current recovery record from the page; it does not immediately delete the previous server files. Their expiry and cleanup still apply.

Shared results

A result is private unless you explicitly create a share link. Anyone who receives that link can request a download without signing in. The link shares the main result, which may include an original DOCX and comments depending on the selected export. Filters applied to a PDF do not remove information from an included original file. Share only with people you intend to give that content to.

Shared links expire with the result. The owning guest session or account can disable a link; this stops new download authorizations. A download URL already issued by Storage can remain usable briefly, and provider caches may delay access changes. Disabling a link does not delete copies someone has already downloaded. The sharing page contains no third-party analytics. Link capabilities are sent in a request body, and our database stores a hash rather than the raw link secret.

Google connection storage and disconnecting

The Google session is stored in an encrypted cookie, marked Secure, HttpOnly and SameSite=Lax. It contains the access token and, if Google provides one, a refresh token. Without a refresh token, the cookie expires with access. With a refresh token, it lasts up to 30 days and can be renewed when access is refreshed. For a server-mode Google import, a separately encrypted copy of the necessary credentials is also stored in the private import record so processing can continue after you close the page. This copy is cleared when acquisition succeeds, fails permanently or is cancelled. Pending copies and document identifiers expire after 24 hours and are then cleared by scheduled cleanup. Temporary failures may be retried before then. Encryption keys are held in server configuration, separate from the database.

A browser-storage flag remembers the connection until you disconnect or clear site storage. These are used for the Google connection you request, not advertising. Selecting “Disconnect Google” clears the cookie and flag and asks Google to revoke access. You can also remove this app in your Google Account. Disconnecting Drive is separate from account sign-out and does not cancel a paid plan or delete files you have downloaded.

To prevent automated guessing of short share links, we temporarily count link checks using a keyed, pseudonymous identifier derived from the request’s network address. These rate-limit records do not store the raw IP address or add the identifier to processing analytics. Counters expire within 24 hours and are removed by scheduled cleanup; deletion may be delayed if cleanup is unavailable. Short-link code hashes remain reserved after expiry so that an old link cannot later point to someone else’s document.

Usage counts and problem reports

The app records daily totals by tool, action and output format. The “Report a problem” form stores a random reference, timestamp, tool and selected problem category. These database records do not contain document content, document links, Google tokens, your email address or a visitor identifier. They help identify broken tools and commonly used formats. They currently have no scheduled automatic deletion.

For server processing, we also keep operational statistics tied to a random document-owner identifier: the selected mode and format categories, whether the source was an upload or a Google import, operation status, timestamps, byte counts, retry counts and stable error codes. These statistics contain no document content, names, document links, author filters, Google tokens, share tokens or IP addresses. Ownership records can link that identifier to an account after sign-in and attachment. Separate usage records track reserved, completed and released exports against the applicable account or guest allowance. We separately count when Storage authorizes a download and when that authorization fails; a successful authorization does not prove that a file was saved to your device. A share recipient's identity is not added to these records. Statistics start when an operation is created on the server; we cannot observe every interrupted browser upload. Operational statistics are scheduled for deletion after 90 days, with bounded cleanup batches; delays are possible if cleanup is unavailable. The underlying processing and allowance records have a separate retention policy.

If you email us, we receive your email address, message and any attachments you choose to send. We use these to answer your request and retain the correspondence for as long as needed to handle it, resolve related issues and meet applicable record-keeping obligations. Please avoid sending documents or personal information that are not needed for support.

Service providers and international processing

The Vercel deployment uses Vercel for hosting and request processing, and Supabase for server-mode file storage, the database, job queue and optional account authentication. Stripe handles enabled payment flows; Resend handles purchase notifications when configured. The original OpenAI Sites deployment uses OpenAI Sites hosting with Cloudflare request-processing and database infrastructure. The providers for the deployment you visit handle connection metadata such as IP addresses, request times, browser information and security logs. Document uploads and results in server mode are handled by Vercel and Supabase as described above.

Google provides Docs, Drive, sign-in and the Gmail mailbox used for support. These providers may process data outside the European Economic Area, including in the United States. Their privacy notices describe their processing, retention criteria and international-transfer arrangements, including adequacy frameworks and contractual safeguards where applicable: Vercel, Supabase, Stripe, Resend, OpenAI, Cloudflare and Google. Provider logs have their own retention rules; connection metadata does not necessarily disappear when you close the page. Contact us for information about the arrangements relevant to your request.

Purposes and legal bases

We process information needed to provide the document tools and related support you request under Article 6(1)(b) GDPR. Our interests in securing the website, investigating failures and understanding aggregate tool usage are the basis for the corresponding processing under Article 6(1)(f). Where a legal obligation requires keeping or disclosing information, Article 6(1)(c) applies. If a separate action relies on consent, you can withdraw that consent without affecting processing already carried out lawfully.

Providing a document or Google access is optional, but the associated import cannot work without it. Google authorization can be revoked at any time. The application does not use personal data for automated decisions with legal or similarly significant effects.

Your rights

Subject to the GDPR's conditions, you can request access, correction, erasure, restriction and portability of your personal data. You can object to processing based on legitimate interests. Email exportmymap@gmail.com with your request; include a report reference if your request concerns a submitted problem report. We may need information to establish which records relate to you and verify your identity.

You can complain to a data-protection authority, including the Berliner Beauftragte für Datenschutz und Informationsfreiheit, or the authority where you live or work. You do not have to contact us first.

Changes

We update this page when the service or its data handling changes. The date above identifies this version.