CMD+RVL sub-processors

This page lists the third-party providers CMD+RVL may use to operate the website, run Cairn, and deliver scoped services. It is written for security, privacy, and procurement review.

Provider use depends on the specific website function, deployment pattern, and customer workflow. For a security package or engagement-specific review, contact drew@cmdrvl.com or use the contact page.

The public website uses Google Analytics 4 for aggregate traffic measurement only after a visitor allows analytics. Every browser request stays on cmdrvl.com: both the analytics script and its measurement beacons are proxied through our own CDN, so the browser never connects to googletagmanager.com or google-analytics.com. This keeps the site rendering inside locked-down corporate networks that block third-party domains. It does not change where the data goes. The measurement data is still processed by Google, which is why Google is listed as a sub-processor above.

Until approval, the Google tag is not loaded and no analytics measurement request is sent. Do Not Track and Global Privacy Control override a stored approval. Advertising storage, advertising user data, advertising personalization, Google Signals, and advertising-personalization signals are disabled. Requests necessarily include network information such as an IP address; Google states that Google Analytics 4 does not log or store individual IP addresses.

Current sub-processors

• Cloud service providers

  Amazon Web Services (AWS), Microsoft Azure

  Role

  Cloud infrastructure, storage, and deployment services.

  Data category

  Application data, operational metadata, and service logs when required for a scoped deployment.

• Identity management providers

  Auth0 by Okta

  Role

  Authentication, authorization, and identity management.

  Data category

  Account identifiers, authentication events, and access-control metadata.

• AI decision providers

  TypeSafe AI, Inc. (Jev)

  Role

  Classification, scoring, and yes/no decisions over short text for platform services, called directly through TypeSafe's API. Jev generates no text.

  Data category

  Catalog metadata such as table names, descriptions, and column lists, and, later, document excerpts. KYC names are never sent to Jev.

• Website analytics providers

  Google (Google Analytics 4)

  Role

  Consent-based aggregate public-website traffic measurement. Served through a first-party proxy on cmdrvl.com, so the browser makes no third-party connection. Not used for advertising or remarketing.

  Data category

  Consented page views, referrer, approximate geography, and device or browser type. The site does not send names, email addresses, account user IDs, or advertising identifiers.

• AI model providers

  Anthropic

  Role

  Model inference for Cairn and for scoped services. We call Anthropic directly through its API, on commercial API terms that do not allow training on the content we send. For a hired engagement, the model configuration is agreed with the client, and zero-data-retention terms are available when the client needs them.

  Data category

  The messages and documents a person sends for a job, and the text needed to answer. Anthropic keeps requests for up to 30 days for safety monitoring, then deletes them; content flagged for a policy violation can be kept longer. Names sent for a KYC check are not sent to any model provider: the sanctions and watchlists are downloaded and matched on our own servers. Text embeddings are computed inside our own containers.

• Messaging providers

  Twilio, Meta (WhatsApp)

  Role

  Text message delivery and phone number verification for Cairn through Twilio. WhatsApp messages pass through Twilio and WhatsApp.

  Data category

  Phone numbers, message content, and delivery status.

• Application database and sign-in providers

  Supabase

  Role

  Database and phone sign-in for Cairn accounts.

  Data category

  Phone number, name, email address, messages, and account settings.

• Email providers

  Resend, Google Workspace

  Role

  Resend sends email from Cairn. Google Workspace receives email sent to CMD+RVL addresses, including cairn@cmdrvl.com.

  Data category

  Email addresses, message content, and attachments.

• Web hosting providers

  Netlify

  Role

  Hosting for textcairn.com, hirecairn.com, and dealcharts.org.

  Data category

  Standard web request logs, such as IP address and browser type.

• Payment providers

  Stripe

  Role

  Card payments for the workshop and the monthly and weekly options. Payment happens on Stripe's own pages.

  Data category

  Name, email address, billing details, and payment card data, which Stripe collects directly. CMD+RVL does not receive card numbers.

• Scheduling providers

  Calendly

  Role

  Booking calls, workshops, and interviews. Booking happens on Calendly's own pages.

  Data category

  Name, email address, the time booked, and anything typed into the booking form.

TypeSafe data use and retention

TypeSafe keeps what we send it only as long as necessary to provide the service; it does not train on it. Zero-retention terms are available under an enterprise agreement. TypeSafe does not publish a fixed retention period for standard API accounts. We call TypeSafe directly through its API. KYC names stay on our own servers and are never sent to Jev or any other AI provider. For paid Hire Cairn engagements, the agreement names the models used.

Provider documents: privacy policy, data processing addendum, enterprise retention information, TypeSafe sub-processors, and security information.

What runs on this site

A complete inventory of every script, form, and integration on cmdrvl.com, what each collects, and where the data goes. This is the whole list; if a flow is not here, it does not exist on the site.

• Analytics

  Google Analytics 4, consent-gated

  Loads only after a visitor selects Allow analytics. Collects page views, referrer, approximate geography, and device type. Off by default, Do Not Track and Global Privacy Control override a stored approval, and advertising storage, Google Signals, and personalization are permanently disabled. Data is processed by Google as the sub-processor listed above.

• Forms

  Contact and lead forms

  The contact page, homepage note, outcome scoping wizard, and PBIX report request collect name, company, email, and a message. Submissions are relayed as email to our inbox through our own AWS infrastructure and are not stored in a database. They are not used for advertising, enrichment, or AI training.

• Uploads

  PBIX auditor files

  Uploaded Power BI files are processed in memory, structural metadata is returned to your browser, and the file is discarded. Report data values are not returned or stored, and no AI model processes the file.

• Everything else

  No other collection exists

  No chat widgets, no advertising pixels, no social trackers, no third-party fonts or CDNs. Every request your browser makes stays on cmdrvl.com.

This inventory is enforced, not just published: an automated browser test runs on every release and fails the build if any page makes a request to a third-party host.

Review notes

• Sub-processor relevance is scoped during security review.
• Customer deployment patterns can change which providers apply.
• Website analytics are separate from scoped customer outcome work.
• Questions and DDQ requests should include the workflow being reviewed.

Sub-processor questions