// Legal

Sub-processors

Third-party services used to operate DraftIQ

Last updated: April 2026 · Version 1.1
What this page covers. Under PIPEDA and applicable privacy law, DraftIQ is required to disclose the third-party services that process personal data or document content on our behalf. This page lists every sub-processor DraftIQ currently uses, what data is shared with them, and why.

If you have questions about how your data is handled, contact hello@draftiq.ca.
ProviderCategoryPurposeData sharedLocationPrivacy policy
Anthropic, Inc.AI processingAI vision model for document extraction. Used when vector parsing is not available (scanned or raster PDFs) and for vision-based recovery passes. Performs visual analysis of rasterized engineering drawings and structures output into valve schedules, metadata registers, and archive registers. Vector symbol detection is performed client-side and does not use this sub-processor.Rasterized page images (when vision fallback is needed)
Extraction prompts
United States
Google Cloud (Vision AI)OCR processingDeterministic text detection (OCR) for engineering drawings. Reads all visible text on P&ID drawings including valve tags, line numbers, spec codes, and pipe sizes. Used as the foundation layer alongside AI vision — provides character-accurate text extraction that cross-validates AI results.Rasterized page images
(processed in real-time, not retained)
United States
Brevo (Sendinblue)Transactional and lifecycle email delivery for welcome messages, access codes, and onboarding sequences.Name
Email address
France / EU
Stripe, Inc.PaymentsSubscription billing and payment processing. Handles checkout sessions, payment methods, and subscription management.Name
Email address
Payment details (handled by Stripe, not stored by DraftIQ)
United States
Netlify, Inc.InfrastructureCloud hosting and serverless function execution (website and backend). User account creation and authentication (Netlify Identity). Session-based usage tracking (Netlify Blobs). Design Partner application form submissions (Netlify Forms).Account registration data
Session tokens
Usage logs (timestamps, file sizes)
Form submission data
United States
(global CDN)

Client-side processing (no sub-processor)

Vector symbol detection — For vector PDFs (CAD-exported drawings), geometric symbol detection and PDF drawing command analysis run entirely in the user's browser. No drawing data is transmitted to any third-party service for this step. Only the structured extraction results are sent to DraftIQ's server for formatting.

Data training

DraftIQ does not use customer documents to train its own models. DraftIQ does not operate AI models and has no model training program.

For third-party AI providers: DraftIQ uses API configurations designed to minimize data retention and prevent model training where supported by vendor terms. Retention and use of data by Anthropic and Google Cloud are ultimately governed by their respective API terms of service. DraftIQ does not opt in to any voluntary data training programs with any sub-processor.

Data retention by sub-processors

Document content sent to Anthropic is processed in real-time and not retained beyond the immediate API request. DraftIQ itself does not store uploaded documents after extraction is complete. Email delivery records are retained by Brevo for standard transactional logging periods. Stripe retains payment records in accordance with financial regulations. Netlify retains hosting logs in accordance with their privacy policy.

Changes to this list

DraftIQ will update this page when sub-processors are added, removed, or materially changed. The "Last updated" date above reflects the most recent change. If you have an active Design Partner agreement, material changes to sub-processors will be communicated by email to hello@draftiq.ca registered addresses before taking effect.

Questions

Contact hello@draftiq.ca with any questions about sub-processor data handling.