Sub-Processor Register

Effective date: 2026-05-10  ·  Version: 1.0  ·  Last updated: 2026-05-10


Purpose

This register exists for three audiences:

  1. Subscribers — to know exactly which third parties have access to their personal data and where that data is stored.
  2. Mexican regulators (INAI) — as the operative documentation supporting our compliance with LFPDPPP Article 36 (international transfer) and the responsable's general duty to maintain a record of processing activities under Mexican data-protection law.
  3. Counsel and auditors — to verify that contractual safeguards are in place with each sub-processor and that the documented purposes match actual data flows.

We (Jaime Muriel Monroy) act as responsable del tratamiento (data controller) under LFPDPPP. The sub-processors below act as encargados (data processors) and process personal data only on our documented instructions.

Current sub-processors

As of the effective date, the following sub-processors handle personal data on our behalf:

1. Clerk

Legal entityClerk Inc.
ServiceUser authentication and session management
Personal data processedEmail, name, optional profile photo, hashed password (when password auth is used), OAuth tokens, IP address, user-agent, session timestamps
Storage locationUnited States (Clerk's primary data centers)
Transfer mechanismStandard contractual safeguards in Clerk's DPA, equivalent to LFPDPPP Article 36 "principio de protección suficiente"
DPA referenceClerk Customer Terms of Service + Data Processing Addendum, accepted in connection with Dardo's account onboarding
Sub-sub-processorsClerk's documented sub-processors as listed at clerk.com/legal/dpa
RetentionPer Dardo's instructions: deleted on user account closure plus 12-month grace, except for items required by law (audit logs)
CriticalYes — without Clerk, users cannot sign in

2. Stripe

Legal entityStripe Payments Mexico, S. de R.L. de C.V. (Mexican merchant-of-record functions); Stripe Payments Company (United States, transaction processing)
ServicePayment processing, subscription billing, customer billing portal, invoicing
Personal data processedName, email, billing address, last four digits of payment card, payment-method type, subscription history, invoice records. Full card numbers, CVV, and expiry are never seen by Dardo — they are tokenized inside Stripe's vault.
Storage locationUnited States primarily (Stripe Payments Company); Mexican entity for local merchant-of-record functions
Transfer mechanismStripe's Data Processing Agreement, with US-transfer safeguards equivalent to LFPDPPP Article 36 "principio de protección suficiente"
DPA referenceStripe Services Agreement + Data Processing Addendum, accepted in connection with Dardo's account onboarding
Sub-sub-processorsPer Stripe's published list at stripe.com/legal/privacy-center
RetentionStripe retains transaction records for at least 5 years per their financial-services obligations; Dardo retains invoice metadata for the same period per Código Fiscal de la Federación Article 30
CriticalYes — without Stripe, no paid subscription can be charged

3. Resend

Legal entityResend, Inc.
ServiceTransactional email delivery (welcome, trial-expiring, payment-failed, subscription-confirmed)
Personal data processedEmail address, name, transactional email content (subject + body), delivery / open / click telemetry
Storage locationUnited States primarily (Resend's primary region)
Transfer mechanismResend's DPA with contractual safeguards equivalent to LFPDPPP Article 36 "principio de protección suficiente"
DPA referenceResend Master Services Agreement + Data Processing Addendum, accepted in connection with Dardo's account onboarding
Sub-sub-processorsPer Resend's published sub-processor list (current at resend.com/legal)
RetentionEmail-delivery telemetry retained 90 days then aggregated; email content not stored after delivery
CriticalOperationally important but not critical — if Resend fails, transactional notifications are delayed but the Service remains operable

4. Cloudflare

Legal entityCloudflare, Inc.
Service(a) Web hosting via Cloudflare Pages for the dardoresearch.com frontend; (b) Worker compute for the API backend; (c) D1 (SQLite) for user records and subscription state; (d) KV for content blob storage; (e) edge log retention for security and abuse detection
Personal data processedIP address, user-agent, request paths, timestamps, session cookies (set by Clerk and traversing Cloudflare edge); plus user records in D1 (email, name, tier, subscription dates)
Storage locationGlobal edge — D1 stores in a primary write region assigned by Cloudflare based on first-write location (no Mexican data center is designated as primary), with read replicas at edge locations; Workers and Pages execute at the user's nearest edge POP. KV is replicated globally. Data may transit through US, EU, or other Cloudflare locations depending on routing
Transfer mechanismCloudflare's Customer Terms + Data Processing Agreement, with cross-border transfer safeguards equivalent to LFPDPPP Article 36 "principio de protección suficiente"
DPA referenceCloudflare Customer Terms + Data Processing Addendum, accepted in connection with Dardo's account onboarding
Sub-sub-processorsNone — Cloudflare operates its own infrastructure; documented at cloudflare.com/cloudflare-customer-subprocessors
RetentionEdge access logs retained 30 days then aggregated; D1 records retained per Dardo's deletion policy in Privacy §7; KV blobs retained per Dardo's content-management policy
CriticalYes — the Service runs on Cloudflare; no fallback hosting

5. Anthropic

Legal entityAnthropic, PBC
ServiceGenerative AI inference (Claude) for daily market briefs, macro reports, and RV signal commentary
Personal data processedNone — Anthropic processes only anonymized market analytics (yield curves, signal z-scores, regime states). User identity and subscription details are never sent to Anthropic. The brief/macro generation pipeline is one-shot inference per cron fire and does not persist user-level data.
Storage locationUnited States (Anthropic's API infrastructure)
Transfer mechanismAnthropic Commercial Terms; personal data is not transmitted to Anthropic in any case (the Service sends only anonymized market analytics to Anthropic for brief generation)
DPA referenceAnthropic Commercial Terms of Service, accepted in connection with Dardo's account onboarding
Sub-sub-processorsAnthropic's documented infrastructure providers; current list available at anthropic.com/legal
RetentionZero — API calls are not retained beyond the inference response per zero-retention mode
CriticalOperationally important but not strictly personal-data critical — Anthropic does not see personal data; if Anthropic is unavailable, daily briefs/macros are not generated but the Service's analytical surfaces remain operable

Sub-processors NOT used

For clarity and to forestall future questions, the following common SaaS sub-processors are not used by Dardo:

The Service explicitly avoids these to minimize data exposure and simplify regulatory posture. Adding any such processor in the future would require updating this register, providing 30-day notice per Privacy §12, and contractual review.

Adding or changing sub-processors

A new sub-processor or material change in an existing sub-processor's scope is a material change to data processing and triggers:

  1. Internal review — verify a current DPA exists or is executable, that the processor's data-protection commitments meet LFPDPPP Article 36 standards, and that the purpose is necessary.
  2. 30-day advance notice to all active subscribers via email and an in-Service notice, per Privacy Policy §12.
  3. Update to this register — bump version, add an entry, update the effective date.
  4. Right of objection — subscribers who object to a new sub-processor may cancel their subscription; the change does not retroactively affect data already processed under the prior register.

Emergency changes (e.g., a sub-processor experiencing a security incident requiring immediate replacement) may be made without 30-day notice but require notice within 30 days after the change and full disclosure of the rationale.

Sub-processor data flow summary

The end-to-end flow for personal data through these sub-processors:

[Subscriber] ─sign-up─▶ [Clerk]              (auth, identity)
      │                    │
      │                    └─webhook─▶ [Cloudflare D1]   (Dardo user record)
      │
      ├─subscribe─▶ [Stripe]            (billing, payment method)
      │                  │
      │                  └─webhook─▶ [Cloudflare D1]   (subscription state)
      │
      ├─uses Service─▶ [Cloudflare Pages + Workers + KV]
      │
      └─receives email─▶ [Resend]      (transactional only)

[Cron]─▶ [Anthropic]   (anonymized analytics in, brief content out)
         │
         └─writes─▶ [Cloudflare KV]    (content blobs read by all subscribers)

No personal data flows to Anthropic. Anthropic receives only aggregated market analytics produced by Dardo's pipeline.

Audit and verification

DPAs with each sub-processor are maintained in Dardo's internal records and are available for INAI inspection on reasonable notice. Each DPA is reviewed at least annually for material changes by the sub-processor and re-executed if required.

A copy of the most recent version of this register is maintained at dardoresearch.com/sub-processors for subscriber reference.

Document v1.0 published 2026-05-10. Companion to Privacy Policy v1.0. Next review: 12 months from publish, on any sub-processor change, or sooner if a material change is required.