Security & operational safety

Only named members of our team get access to your systems, never a shared or anonymous account. Access is scoped to what the project actually needs, nothing broader. You can revoke it at any time, and we remove it as a standard step once a project is handed over.

API keys and other credentials are stored in the credential management system of whichever automation platform we're using for your project, not in spreadsheets, chat messages, or workflow code. Credentials are never hardcoded into a workflow and never sent to you or anyone else over email or messenger.

Some actions shouldn't run unattended: sending an invoice, issuing a refund, deleting a record. Where an action is sensitive or hard to reverse, we build in an approval step so a person confirms it before it happens. Before any automation goes live, we test it against real data, apply sensible rate limits, and add safeguards against duplicate actions.

Automations include logging and error handling appropriate to what they do, so when something breaks, we know about it, often before you do. Where it matters, failures trigger an alert to our team rather than failing silently. This is part of the ongoing monitoring included in every engagement.

Every project includes a data processing agreement (DPA) on request, setting out what data we process, why, and for how long. Hosting in the EU is available as an option for clients who need it; it isn't a blanket guarantee for every tool we integrate with, since some third-party services you already use may host data outside the EU under their own terms. The subprocessors we use for the site and client communication are listed in our Privacy Policy.

Once your project is paid for, the workflows, configurations, and documentation we build are yours. If you ever want to move on, we hand everything over, credentials removed from our side, documentation included, so you're not locked into working with us.