Bulk extract data from PDFs

PDF BulkX is a cloud service for bulk extraction of structured data from PDF files. Upload your PDFs in bulk — invoices, receipts, or reports — and receive clean Excel or CSV output in seconds.

Logic Mode learns your PDF layout once during setup, then applies the same logic to every file to get stable results across files of the same type. AI Extract Mode reads each PDF with Google Gemini on every run; it handles scanned documents and layouts that vary from file to file, with AI flexibility.

AI Extract Mode runs Gemini against fixed columns you define once at project setup. Every extraction run returns results in your defined columns, in the order of uploaded files — no need to write a new prompt for each file, no need to strip Gemini's lengthy explanations from its replies. You also get automatic retries when Gemini fails with no extra charge, monthly credit tracking, and during setup a side-by-side preview comparing extracted data against sample PDFs so you can verify accuracy before running bulk extractions — none of which calling Gemini directly provides.

Both Logic Mode and AI Extract Mode use **1 credit per PDF page** — a 10-page PDF costs 10 credits in either mode. Each tier comes with a monthly credit allowance, and you can buy additional top-up packs anytime. Monthly credits are consumed first; top-up credits are used only after monthly credits run out. After sign-in, your current balance is visible in the top navigation bar on every page, with the full monthly/top-up breakdown on the billing page.

Top-up packs are one-time credit purchases that add to your account on top of your monthly tier credits. Top-up credits **never expire** while your account is active and carry over indefinitely — they are only forfeited if you permanently delete your account. When you run extractions, monthly tier credits are consumed first. Top-up credits are used only after the monthly allowance for the period runs out, so a top-up pack effectively extends a heavy month rather than replacing your subscription. Top-up purchases are **non-refundable** once the credits are granted, since you can immediately use them. Duplicate charges or billing errors are reviewed case-by-case — contact support@expertsmix.com if you need assistance.

Both digital (machine-readable, text-selectable) PDFs and scanned PDFs are supported. Logic Mode works best with digital PDFs. AI Extract Mode can handle scanned documents and varied layouts.

For digital PDF files with a consistent layout, Logic Mode typically achieves near-100% accuracy after setup. AI Extract Mode accuracy depends on document quality and Google Gemini's capability for your document type. You can review and refine results before running bulk extractions — the system shows extracted data alongside the original PDF for easy comparison.

Yes. The Free tier lets you create 1 project (one document type) with 8 credits per month, including both Logic Mode and AI Extract Mode. No credit card required.

We accept major credit and debit cards (Visa, Mastercard, Amex) for subscriptions. For one-time top-up packs, you can pay with card or PromptPay. Subscriptions currently require a card. All payments are processed securely by Stripe.

**Cancel anytime via your account settings** — go to Settings → Plan & Billing → Manage Subscription, then choose "Cancel subscription". You'll keep access until the end of the current billing period; no further renewals are charged. **Within 7 days of subscription start**, you can request a pro-rated refund under Thailand's Direct Sales and Direct Marketing Act §33. Email support@expertsmix.com — we process refunds per the formula in our Terms (Payment & Credits section): refund = original payment − the larger of (a) the value of credits already consumed, or (b) the time-based pro-rate of the billing period. **After the 7-day window**, cancellation takes effect at the end of the current billing period and no refund is issued for the remaining time. This matches standard practice for subscription services. **Top-up credit packs are non-refundable** once the credits are granted, since they're available for immediate use.

Yes — but only on **annual** subscriptions. Go to Settings → Pause Account. When you resume, your billing period end date is extended by the pause duration so no paid time is lost. While paused, no extractions can run and no billing accrues. Monthly subscriptions can't be paused — to stop processing, cancel at the end of the current billing period (Settings → Plan & Billing → Manage Subscription → Cancel subscription).

Go to Settings → Danger Zone → Delete Account. Your account enters a **15-day soft-delete grace period** during which you can cancel the deletion. After 15 days the account, projects, files, and personal data are erased from our active systems. If you have an active subscription, it's automatically scheduled to cancel at the end of the current billing period when you start deletion — no further charges occur. **Top-up credits expire** at the moment of permanent deletion (the 15-day grace lets you change your mind to keep them). A minimal record (email, name, account creation timestamp) is retained in our audit log for 365 days after permanent deletion, as required by Section 26 paragraph 2 of Thailand's Computer Crime Act. See the Privacy Policy for full retention details.

PDF BulkX is currently below Thailand's VAT registration threshold (฿1.8M annual revenue), so we do not yet issue a full Thai tax invoice (ใบกำกับภาษี) with VAT line. You will receive a branded receipt (ใบเสร็จรับเงิน) by email after each completed payment, suitable for expense records. We will offer full Thai tax invoice once we register for VAT.

PDF BulkX is PDPA-aligned. Sample PDFs uploaded during project setup are kept for up to 7 days after you save (the grace window lets you refine extraction logic), then auto-deleted. PDFs uploaded for bulk extraction (within a saved project) are deleted immediately after each run completes. PDFs uploaded via Quick Mode are kept for up to 24 hours, then auto-deleted unless you save the extraction as a project. Extracted results (Excel/CSV) are deleted 24 hours after the run completes. No data is used to train AI models. A complete audit log of all data-processing events is kept for 365 days. You can export or delete all your data at any time from account settings.

Top-up credits (purchased separately) never expire while your account is active and carry over indefinitely. They expire upon permanent account deletion (15-day soft-delete grace period followed by hard delete). Monthly tier credits reset at each billing cycle and do not carry over. If you downgrade, you keep your current monthly credits until the end of the current billing period.

AI processing uses Google Gemini in the United States, with explicit consent and PDPA-aligned safeguards. Other data processing happens on our infrastructure in Singapore. File storage uses Cloudflare R2 in the APAC region. Full details, including countries and sub-processors, are in our Privacy Policy.

Yes — PDF BulkX Enterprise is a self-hosted version that runs on your own servers, with custom integration options for your environment. Best for organizations with infrastructure ownership requirements, larger volumes than Cloud tiers support, or specific compliance needs. Learn more at https://expertsmix.com/products/pdf-bulkx

AI Extract Mode reads documents in most languages Google Gemini supports — Thai, English, Chinese, Japanese, and 100+ more. Logic Mode also works with most languages — it learns where data sits on the page, not what language the text is in. As long as the document structure stays consistent, language doesn't matter. The web interface is currently in English. Thai version coming soon.

Yes. During project setup and for up to 7 days afterward, you can iterate on natural-language refinement prompts — for example, reformat dates, merge columns, or skip certain rows. After 7 days, sample PDFs are deleted (per PDPA §37(3)), and to refine the logic further you re-upload sample PDFs and run setup again. Your extraction setup is preserved as a version history across runs.