Frequently Asked Questions

Last updated: 7 November 2025

Do you store my PDFs?

We process files in memory and short-lived storage only to produce results. We do not persist original PDFs after processing completes, except transiently to deliver downloads.

Limited metadata (e.g., session IDs, timestamps, and error reports) is kept for fraud/abuse prevention and automatically cleaned up after 30 days by the anonymous session cleanup job. Extracted JSON/CSV snippets are retained only as needed to render downloads.

Are my files used to train AI?

When using Google Gemini (Paid Services) and OpenAI API, providers state API content is not used for model training by default. They may retain limited logs briefly for safety and legal compliance.

Where is data processed?

We aim to run in EU regions where possible (hosting, DB). Some providers may process data outside the EU/CH/UK with appropriate safeguards; see /legal/subprocessors.

We update the Subprocessors list whenever a new provider or region is added so you can verify the applicable safeguards, DPAs, and SCCs. Data residency controls are applied in the Supabase + Vercel configs whenever available.

Is pdfextractorai GDPR/FADP compliant?

We designed the Service with privacy by default, provide a DPA where applicable via our providers, and support user rights (access, deletion, portability). See Privacy Policy for details.

European and Swiss visitors can view their consent choices, download a data export, or submit deletion requests via the Privacy Policy controls. We track consent history, log-processing activities, and provide records for compliance purposes.

What data do you collect?

We collect account/auth data (if you register), limited usage diagnostics, and the documents you upload during extraction. Uploaded PDFs are processed transiently, and payment data is handled by Stripe when billing is enabled.

A full breakdown of categories, retention, and legal bases is on the Privacy Policy.

Where can I review the Terms of Service?

The Terms cover acceptable use, AI accuracy expectations, age requirements (18+), availability, and billing terms.

Read the full terms at /terms.

Which subprocessors handle data?

We rely on Vercel for hosting, Supabase for auth/database, and AI providers such as Google Gemini or OpenAI. Stripe and email providers are added when paid plans and notifications are enabled.

The current list, locations, and safeguards are detailed on /legal/subprocessors.

What about accuracy?

AI extraction can be imperfect. You must review and verify results before relying on them for accounting or other decisions.

Can I delete my data?

Yes. Delete your account in-app (if available) or email privacy@pdfextractorai.com. We'll delete account data not required by law (e.g., invoices).

Anonymous sessions are automatically purged after 30 days by the cleanup job. Signed-in users can request deletions through the GDPR controls on the privacy page. We log deletion requests for audit and reply via the contact email listed above.

Do you block certain regions?

We may apply IP-based geoblocking/rate-limiting for abuse prevention or phased launches. IP addresses are personal data and used for security/legitimate interests.

What's the age requirement?

The Service is for users 18+.

Do you support OpenAI or only Gemini?

We currently support Gemini; OpenAI may be available as an option. See Privacy for provider details.

How many extractions can I run anonymously?

Anonymous visitors are limited to 3 extractions per day, tracked by a session identifier stored in session storage. Each session expires after 24 hours, and all anonymous usage records are purged after 30 days via the cleanup job.

We also enforce per-IP caps through the `/api/anonymous/session` endpoint to prevent abuse; once the quota is reached, you will receive a 429 response until the reset time.

How can I reach you?

Visit the Contact page for the latest support channels. Until our Resend-verified domain email addresses are ready, please use contact@pdfextractorai.com.

Once the domain and professional mailboxes are live, we will swap the placeholder addresses for support@pdfextractorai.com, tech@pdfextractorai.com, and the business inbox you see on the contact page.