Security and data handling
Your queries pass through. Nothing stays.
What your security, legal and finance teams ask before they sign off: what we keep, what we never keep, who we are and how to buy. Anything not answered here, send it to us and we answer in writing.
Your requests
- Queries are processed in memory
- Your queries, request parameters and MCP context are processed in memory for the duration of each request. Once a successful response is delivered, no trace of your request payload remains on our systems.
- One exception, stated plainly
- When a request fails with an error on our side, its parameters go to our own diagnostic logs so we can reproduce and fix it. Those logs stay on infrastructure we operate, are not archived or backed up, are not sent to any third party, and are destroyed whenever the service is next deployed.
- Results are not stored
- The data we return passes through to you. We do not store, cache, index or archive it, and every request is answered live.
- What each API call records
- For billing and abuse prevention we record your user ID, API key ID, endpoint name, HTTP status code, response latency and timestamp. Never your query parameters, request body or response content. Privacy policy
Company and GDPR
- An EU company
- FetchLayer is run by HootCodes LTD, registered in Sofia, Bulgaria. Our supervisory authority is the Commission for Personal Data Protection (CPDP). Company details
- Data protection requests
- GDPR and other data protection requests go to privacy@hoot.codes. Where the GDPR applies, we process account data to provide the service, to keep it secure, and to meet tax and accounting duties.
- Your role with public data
- Public content can still contain personal data. When your product processes it, you decide whether that use is permitted and act as the controller for it, as our privacy policy sets out.
Service providers
- Hosting and infrastructure
- Hosting and infrastructure providers in the EU and US; details available to customers on request under NDA.
- Payments and email
- Polar processes payments and issues invoices. Postmark sends transactional email, such as sign-in codes and account notices.
- AI chat
- OpenRouter provides the language models behind Research Chat, and only receives what you send in that chat.
- Sign-in and scheduling
- Google and GitHub, only when you choose to sign in with them. Calendly, only when you book a call with us.
Retention and security
- Account and billing records
- Kept while your account is active, and afterwards for as long as legal, accounting and dispute purposes require.
- Sign-in and sign-up records
- Sign-in verification records expire quickly. At sign-up we keep one-way, keyed hashes of the IP address, network and browser for up to 90 days, only to stop repeated use of free trials. They cannot be turned back into the original values.
- Security measures
- Access controls, rate limiting and logging, with every key authenticating each request. No system is perfectly secure, and we do not claim one is.
Procurement
- A signed terms letter
- Your price, the guarantee and the ceiling after it, plus our uptime commitment, in a signed letter you can verify online. The plan itself is month to month. Our SLA
- Invoices
- One subscription, paid by card through our checkout, with a downloadable invoice for every payment in your billing portal. Tax is added at checkout where it applies.
- Payment terms
- If your finance team needs a different way to pay, tell us what it needs and we will look at it with you.
- Your forms and questionnaires
- Send us your MSA, DPA, vendor forms or security questionnaire, and we will go through them with you and answer in writing.
Common questions.
Do you store our queries?
No. Queries and request parameters are processed in memory, and nothing of your request payload remains once a successful response is delivered. The one exception is a request that fails with an error on our side: its parameters go to our own short-lived diagnostic logs so we can fix it.
Do you keep the data you return?
No. Results pass through to you and are not stored, cached, indexed or archived. Every request is answered live.
What do you log about each API call?
Your user ID, API key ID, endpoint name, HTTP status code, response latency and timestamp, for billing and abuse prevention. Never your query parameters, request body or response content.
Is FetchLayer GDPR compliant?
We are an EU company, HootCodes LTD in Bulgaria, and the GDPR applies to how we handle your account data. Data protection requests go to privacy@hoot.codes, and our privacy policy sets out the legal bases and retention.
Which service providers process our account data?
Payments: Polar. Email: Postmark. AI chat: OpenRouter, for Research Chat only. Sign-in: Google and GitHub when you choose them. Call booking: Calendly. Hosting and infrastructure providers in the EU and US; details available to customers on request under NDA.
Will you sign our DPA or fill in our security questionnaire?
Send them over and we will go through them with you and answer in writing.
Questions from your review?
Send them in a call or by chat, and we answer in writing. Our uptime commitment is on the SLA page.