Effective Date: [Date] Company: [Your LLC Name], a California Limited Liability Company
1. Introduction
[Your Company Name] (“we,” “us,” or “our”) respects your privacy and is committed to protecting the sensitive information you entrust to us. This Privacy Policy specifically governs our Bulk PDF-to-Markdown Processing Service. It describes how we handle, process, and destroy the documents and data you submit to us.
Key Privacy Commitment: We utilize self-hosted, open-source Artificial Intelligence models. We do not use your data to train foundation models, nor do we share your data with third-party AI providers (such as OpenAI or Anthropic) for their improvement purposes.
2. Information We Collect
We collect two distinct types of information:
A. Client Business Information (Transactional)
To facilitate the service, we collect your name, company name, email address, phone number, and specific job requirements.
Payment Information: We use Stripe for payment processing. When you pay for a job, you are directed to Stripe’s secure hosted checkout. We do not collect, see, or store your credit card number, bank account details, or CVC. We only receive a confirmation transaction ID and payment status from Stripe.
B. Service Data (The Files)
This includes the actual PDF documents you transfer to us for processing. We process this data strictly for the purpose of converting it to the requested format (Markdown/Image Descriptions).
3. The AI Processing Architecture (How We Handle Your Files)
We have designed our technical architecture to maximize security and data isolation.
- Infrastructure: We utilize ephemeral (temporary) Virtual Machines (VMs) hosted in secure Data Centers located within the United States.
- The AI Model: We run open-source Optical Character Recognition (OCR) and Large Language Models (LLMs) using vLLM. These models are hosted locally within our private environment.
- No External API Calls: Your document text is not sent to external AI APIs (e.g., ChatGPT, Claude, Gemini). All inference happens within the isolated VM we control.
- Ephemeral Environments: Once a processing job is complete and the VM is no longer needed, the virtual environment is terminated and wiped.
4. Data Retention & Human Review
A. Retention Lifecycle
We operate on a “Process and Purge” basis:
- Ingestion: Files are received via a secure file transfer method (see Section 5).
- Processing: Files are loaded into the VM for conversion.
- Delivery: The converted output is sent to you for review.
- Deletion: Upon your confirmation of receipt (or within 7 days of delivery if no confirmation is received), the source PDFs and the output files are permanently deleted from our local storage and cloud environments.
B. Human Access
Unlike fully automated SaaS platforms, our service includes Quality Assurance (QA). Authorized members of our team may view your files to:
- Verify the accuracy of the markdown conversion.
- Ensure tables and image descriptions are correctly formatted.
- Debug processing errors.
All team members with access are bound by strict confidentiality agreements.
5. Data Sharing & Sub-Processors
We do not sell your data. We only share data with specific infrastructure providers required to deliver the service:
| Provider Type | Purpose | Note |
|---|---|---|
| Cloud Compute Providers | GPU rental & Virtual Machines | We use major US-based providers. Data exists here only during active processing. |
| Stripe | Payment Processing | We share only necessary billing details (Name/Email). |
| Client-Designated Transfer Platforms | File Ingestion & Delivery | To accommodate your security preferences, we may use file transfer platforms designated by you (e.g., your corporate Box, Dropbox, SharePoint, or Google Drive links). In these instances, we access the data solely to download it for processing and upload the results. |
6. Client Warranties & Acceptable Use
Since we manually vet jobs, we reserve the right to refuse service. By submitting data to us, you warrant that:
- You own the data or have the necessary rights/licenses to process it.
- You are not submitting non-public data belonging to a third party (e.g., a competitor) without their consent.
- The data does not contain illegal content.
7. Security Measures
We implement industry-standard security measures, including:
- Encryption: Data is encrypted in transit (TLS/SSL) and at rest where applicable.
- Access Control: Only authorized personnel working on your specific job have access to your files.
- Minimal Logs: We do not retain logs of the content of your files. We only retain communication logs (email) regarding the job scope.
8. Your Rights (California & US)
As a California-based company, we comply with the CCPA/CPRA. Regardless of your location, we extend these rights to all clients:
- Right to Know: You may ask what personal data we have collected (usually limited to your contact info).
- Right to Delete: You may request the deletion of your business contact info (Client Data), subject to our legal tax/accounting retention obligations.
- Right to Opt-Out: You may opt out of marketing communications at any time.
9. Changes to This Policy
We may update this policy as our infrastructure evolves. If we make material changes—specifically regarding how your data is processed or AI model usage—we will notify you via email prior to processing your next job.
10. Contact Us
If you have questions about our security or privacy practices, please contact:
[Your Company Name] [Address/PO Box] [Email Address]
Last Updated: [Date]
1. The Service [Your Company Name] (“Provider,” “we,” or “us”) provides AI-powered bulk PDF-to-Markdown conversion, image description, and table preservation services (the “Service”).
2. Nature of Service & Human Review You acknowledge that this is a Consulting Service, not a fully automated SaaS platform.
- Human-in-the-Loop: Our team members will view your files to perform quality assurance and debugging.
- Job Acceptance: We reserve the right to reject any job for any reason. If we reject a job after payment but before processing begins, a full refund will be issued.
3. User Representations & Warranties By submitting files to us, you represent and warrant that:
- Ownership: You own the copyright to the files or have full legal authority/licenses to process them.
- No Stolen Data: You are not submitting non-public, proprietary data belonging to a third party (including competitors) without their express written consent.
- No Illegal Content: The files do not contain illegal material, hate speech, or content that violates US law.
4. Intellectual Property
- Your Data: You retain all rights, title, and interest in the documents you submit and the converted output we deliver. We claim no ownership over your content.
- No Training: We guarantee that your data is not used to train our AI models or third-party foundation models.
5. Confidentiality We agree to keep your files confidential. We will not share, show, or distribute your files to any third party other than the infrastructure providers listed in our Privacy Policy (e.g., Cloud Compute, Stripe, File Transfer Services) strictly for the purpose of processing.
6. Data Retention We are a “Stateless” processor.
- Once you confirm receipt of your converted files, we delete both the source inputs and the outputs from our systems.
- If we do not hear from you within seven (7) days of delivery, we will automatically delete the files to ensure data hygiene.
- It is your responsibility to download and back up your converted files immediately upon delivery.
7. Disclaimers
- Accuracy: While our pipeline utilizes advanced open-source models, OCR and AI descriptions are not infallible. We provide the converted text “as-is” and do not guarantee 100% accuracy on every character, table cell, or image description.
- Formatting: Complex layouts may require manual adjustment by you after delivery.
8. Limitation of Liability To the maximum extent permitted by California law, [Your Company Name] shall not be liable for any indirect, incidental, special, or consequential damages. Our total liability for any claim arising out of these terms is limited to the amount you paid us for the specific job giving rise to the claim.
9. Indemnification You agree to indemnify, defend, and hold [Your Company Name] harmless from any claims, damages, losses, or legal fees arising from your violation of Section 3 (Representations & Warranties)—specifically, if a third party takes legal action against us claiming you submitted their proprietary data without permission.
10. Entire Agreement & Custom Contracts These Terms constitute the entire agreement between us regarding standard jobs. However, if we have signed a separate Master Services Agreement (MSA) or Statement of Work (SOW) with you, the terms of that signed agreement shall control and supersede these Terms.
11. Governing Law & Dispute Resolution These terms are governed by the laws of the State of California. Any disputes arising from these terms shall be resolved in the courts located within [Your County], California.
