Security and data handling
Where your data goes, and who else touches it
The whole list, including the parts that are not ours. We hold no certification and do not claim one.
We make no accuracy, security or uptime certification. Any certification claimed here must be one the owner actually holds and can evidence.
The data flow, end to end
-
Browser
Your browser talks to our application over HTTPS when you sign in, configure a workflow, or view results.
-
Cloudflare Workers and D1 (EU-West)
Our application runs on Cloudflare Workers, with data stored in Cloudflare D1, a SQL database, in the EU-West region. This is where your workspace configuration, source events and retained evidence live.
-
HubSpot
We read your contact records back from HubSpot, using the read access you grant us, to check them against your rules.
-
Resend
We read message status events back from your connected Resend account to check whether the acknowledgement email reached the point your rules require. Resend is also the service we use to send our own transactional emails to you: sign-in links, failure notices, deletion confirmations.
-
Stripe
Billing is handled entirely by Stripe hosted Checkout and the Stripe Billing Portal. Card details never reach our servers.
-
Optional model provider
Only if you turn on the optional AI assistant: the specific text needed to answer your request is sent to a model provider. The assistant is off by default and never decides a verification result, an access right, or a charge.
Every call we can make with your credentials
This is the whole list. The connectors build every request from a fixed table and cannot construct one that is not in it, so there is no path here that writes to your CRM or sends mail.
| Provider | Call | What it is for |
|---|---|---|
| HubSpot | POST /oauth/v2/private-apps/get/access-token-info | checks the key is live and reads which scopes it carries |
| HubSpot | GET /crm/v3/objects/contacts/ | reads back the one contact an enquiry names |
| HubSpot | POST /crm/v3/objects/contacts/search | finds that contact by your correlation property when no id was given |
| HubSpot | POST /crm/v3/objects/contacts/search | searches contacts that carry your correlation property, by name or email, when you are choosing one to test; it reads names, emails and ids only |
| Resend | GET /emails/ | reads the delivery outcome of one acknowledgement we were told about |
| Resend | GET /domains | checks the key is live, and which sending domains it can see |
| Resend | GET /emails | lists recent messages you have sent, so you can pick one to test rather than hunt for its id |
Subprocessors
| Subprocessor | Role | Data involved |
|---|---|---|
| Cloudflare | Application hosting, database (D1) and cron scheduling | All workspace configuration, source events and retained evidence |
| HubSpot | Customer-connected CRM evidence source (read access only) | The specific contact record fields your workflow rules reference |
| Resend | Customer-connected email evidence source (read access only), and our own transactional email sender | Message status events for the acknowledgement email your workflow rules reference; the transactional messages we send you (sign-in links, failure notices, deletion confirmations) |
| Stripe | Payment processing and billing management | Billing details and subscription status; we never see full card numbers |
| Optional model provider | Powers the optional AI assistant, only when a customer enables it | Only the text needed to answer the specific assistant request |
Retention and access
Evidence we retrieve from HubSpot and Resend is kept for 30 days by default, then removed. This retention period is a fixed system limit, not a per-customer setting.
Do you store my customer data?
We store the specific pieces of evidence needed to check your rules, the CRM record fields and email status your workflow's assertions reference, and the source event your automation sent us, scoped to your workspace. We do not import or mirror your whole CRM. Evidence is kept for 30 days by default and then removed.
Do you ever modify my CRM or resend my emails?
No. We only ever read. We do not create or edit CRM records, and we do not send a replacement email on your behalf. If a run fails, fixing it is still on you.
Is my data used to train anything?
No. Evidence we retrieve is used only to check your rules and show you results. If you turn on the optional AI assistant, only what is needed to answer your specific question is sent to the model provider for that request; the assistant can never decide a verification result, an access right, or a charge, and by default the assistant is switched off.
How long do you keep evidence?
30 days by default, then it is removed.