Sub-processors
Last updated: September 6, 2026
Traceten engages a limited number of third-party service providers ("sub-processors") to operate the analytics service. Active providers are listed first, followed by any planned providers listed ahead of launch. Each provider is contractually bound to security obligations no less protective than those in our Data Processing Agreement.
We notify customers at least 30 days before adding a new sub-processor that processes personal data, by updating this page and emailing the primary administrative contact on the customer's account.
To subscribe to sub-processor change notifications, email legal@traceten.com with the subject "Sub-processor notifications" and the email address you would like added.
Active sub-processors
These providers are live and process personal data on our behalf today. Engaging them is authorised under Section 8 of the Data Processing Agreement.
| Provider | Purpose | Location | Data terms |
|---|---|---|---|
| Cloudflare, Inc. | Edge ingestion (Workers), DDoS protection, IP hashing at the edge. | Global edge; primary entity United States. | DPA |
| ClickHouse, Inc. (ClickHouse Cloud) | Analytics event storage and query. Retention varies by record type, from 90 days for the shortest-lived records to the life of the account for aggregated daily rollups. The Privacy Policy states each period and is the canonical statement. | United States. | DPA |
| Neon (Databricks, Inc.) | Postgres metadata database, including accounts, sites, integrations, API and crawl credentials, and the AI crawler registry. | United States. | DPA |
| Upstash, Inc. | Redis cache for sessions and rate-limit counters. | United States. | DPA |
| Redpanda Data, Inc. (Redpanda Cloud) | Event streaming bus between ingestion and the classification engine. | United States. | DPA |
| Clerk, Inc. | Authentication and identity for the customer dashboard, and the system of record for team membership: the name, email address and role of every person you invite to your account, and every pending invitation. Clerk also sends the invitation email on your behalf, which means it processes the address of an invited person before they are a Traceten user at all. | United States. | DPA |
| Stripe, Inc. | Subscription billing and payment processing for Traceten customers. | United States. | DPA |
One flow is disclosed separately from the table above. AI crawler tracking attempts reverse-DNS verification whenever the published-range check does not match and the crawler's claimed vendor supports it (see section 3 of the Privacy Policy). An attempt sends a lookup to Cloudflare's public DNS resolver at cloudflare-dns.com (the endpoint behind 1.1.1.1) and not to the Cloudflare Workers service listed above. The value sent is the address reported to us with the crawl, before anything is verified: the lookup is the test of whether that address belongs to the crawler it claims to be, so it may belong to someone else entirely. It is sent on attempts that are not already answered from a short-lived cache of recent lookups, and on failed attempts as well as successful ones. We name the endpoint here so the flow is visible; we are confirming which of Cloudflare's terms govern it and will update this page with the answer.
Planned sub-processors
The providers below are listed ahead of launch and are not yet processing personal data. Listing them here is advance visibility, not the formal notice: a planned provider is only authorised once it moves to the active list above, and that promotion triggers a fresh 30-day notice and objection window (see Section 8 of the Data Processing Agreement). No planned provider begins processing personal data until that window has run.
| Provider | Purpose | Location | Data terms |
|---|---|---|---|
| Amazon Web Services, Inc. | Planned cold storage of historical events in S3 (Parquet, SSE-KMS encrypted). Not yet in production; listed ahead of launch. | United States. | DPA |
| Fly.io, Inc. (candidate host) | Planned application hosting for the API and classification services. These process events in transit, including hashed IP addresses, approximate location, pseudonymous visitor and session identifiers, and page URLs, plus customer email addresses in transit on payment events before hashing. No customer data is stored on the host. One of Fly.io or Railway will be selected; the other will be removed from this list. | United States, with configurable region placement. | DPA |
| Railway Corp. (candidate host) | Alternative planned application host for the same services and the same categories of data as the Fly.io entry above. One of the two will be selected; the other will be removed from this list. | United States. | DPA |
| Perplexity AI, Inc. | Planned answer-engine measurement for the AI visibility feature. Receives the prompt text the customer wrote, as a chat message to the Sonar API, together with an output-token limit and an approximate country derived from the locale set on that prompt. No account, site, visitor or session identifier is sent, and nothing collected from the customer's website is sent. Not yet in use; listed ahead of launch. | United States. | DPA |
| DataForSEO OU | Planned search-results retrieval for the Google AI Overviews surface of the AI visibility feature. Receives the prompt text the customer wrote as a Google SEARCH QUERY, not as a chat prompt, together with a location code, a language code and a fixed device and operating-system profile, and runs that search against Google from its own infrastructure. The query text therefore also reaches Google LLC as an ordinary web search; see the note below the planned table. No account, site, visitor or session identifier is sent. Not yet in use; listed ahead of launch. | Estonia. | DPA |
| Anthropic, PBC | Planned answer-engine measurement for the AI visibility feature, available only to customers who buy the engine coverage add-on. Receives the prompt text the customer wrote, as a user message to the Anthropic API, together with an output-token limit, an approximate country derived from the locale set on that prompt, and the web search tool capped at three searches per request. Anthropic's API has no equivalent of OpenAI's store setting, so there is no server-side retention flag for Traceten to set on this path. No account, site, visitor or session identifier is sent. Not yet in use; listed ahead of launch. | United States. | DPA |
| OpenAI OpCo, LLC | Planned answer-engine measurement for the AI visibility feature, available only to customers who buy the engine coverage add-on. Receives the prompt text the customer wrote, as the input to the OpenAI Responses API, together with output-token and tool-call limits, a low reasoning-effort setting, an approximate country derived from the locale set on that prompt, and the web search tool. Traceten sends store: false on every call, which under OpenAI's API terms opts that call out of OpenAI's default server-side retention of the request and response. No account, site, visitor or session identifier is sent. Not yet in use; listed ahead of launch. | United States. | DPA |
Two things about the four AI visibility providers are worth stating plainly. The text those providers receive is the prompt the customer wrote (a question such as "best project management tool for agencies") together with a location signal taken from the language and region the customer set on that prompt: an approximate country for the three chat APIs, and a location and language code for the search query. Neither is derived from any person. No visitor data is sent, and nothing we measure about a customer's website is sent: no visitor or session identifier, no cookie, no IP address or hash of one, no page URL, and no account or site identifier. The prompt itself is customer-authored text, and a customer who builds one from our suggestion tool is sending words taken from their own page headings. The feature is not enabled in any Traceten deployment, so no prompt has been sent to any of these four providers, and none of them is processing personal data on our behalf today.
One onward flow is disclosed here rather than as a separate row. The Google AI Overviews surface reaches Google through DataForSEO, which runs the customer's prompt as an ordinary Google search from its own infrastructure and returns the results page. The prompt text therefore reaches Google LLC as a search query, handled by Google as it handles any other web search. Google is DataForSEO's sub-processor rather than ours, which is why it is named here instead of listed above. Traceten does not use Google's Gemini API or any grounding API, and the retention terms attached to those products do not apply to this path.
Data residency
All personal data listed above is processed in the locations named in the tables above. You select a region when you create a site and we record it, but region-based storage routing is not yet in place, so that setting does not currently change where data is held. EU residency is not available yet. An earlier version of this page said we configure the providers above to use their EU regions on request. That was not accurate, and this is the correction.
Transfers out of the EEA, the UK, and Switzerland rely on the EU Standard Contractual Clauses incorporated into our Data Processing Agreement. We will update this page and notify affected accounts before region routing goes live. If EU residency is a requirement for you, contact legal@traceten.com before you rely on it.
Changes to this list
Material changes to this page (addition, removal, or change of role of a sub-processor that processes personal data) are dated and announced as described above. The "Last updated" date at the top of this page reflects the most recent change.