You can trust us with your data.
Handing sensitive data to an outside partner takes trust, and doing AI work with it takes more. We sign an NDA before we start, work inside your own systems, and never share, sell or train models on what’s yours. No fine print, no surprises. Here’s exactly how we protect your business.
Six ways we keep your information yours
Confidentiality isn’t a clause we bury in a contract: it’s how we work from the first conversation.
An NDA before anything
We sign a mutual NDA in writing before we look at a single row of your data. Yours or ours, whatever you prefer.
You own everything
Every dashboard, pipeline and automation we build is yours, in your accounts, documented, with no lock-in.
Least-privilege access
We request only the access a task actually needs, and it is revoked the moment the work is done.
Never used to train AI
Your data is never sold, shared, or used to train a model. Where an AI workflow touches it, we use business tiers with training switched off.
Inside your own systems
Wherever possible we work in your accounts and your warehouse, so your data never leaves your control.
Removed on request
Credentials and any working copies are deleted at the end of the engagement, or the moment you ask.
Locked down by default, start to finish
The same sequence on every engagement, so protecting your data is never an afterthought.
Sign the NDA
A mutual confidentiality agreement is in place before any access is shared.
Scope the access
We agree exactly which systems we need, and request read-only where we can.
Work in your systems
The build lives in your accounts and warehouse. Your data stays with you.
Revoke & hand over
Access is revoked, working copies removed, and everything we built is yours.
The questions people actually ask
No legalese, just how we treat your data and your trust.
Can you sign our NDA?
Yes. We are happy to sign your NDA, or provide our own mutual one, before any systems or data are shared.
Where does our data live?
In your own accounts and warehouse wherever possible. When a working copy is unavoidable, it is access-controlled and deleted at the end.
Do you reuse our data or trade secrets?
Never. Your data, methods and results are used only for your engagement, and are not shared with other clients or used to train models.
How do you handle passwords and keys?
Through your own password manager or scoped service accounts. We never ask for credentials in plain text, and never store them.
Does our data get sent to AI models?
Only where an AI workflow you approved requires it, and only the fields that workflow needs. We use business and enterprise API tiers where provider training on your data is off by default, we agree the scope in writing first, and for sensitive work we can keep processing inside your own cloud account.
You build with AI. Who is actually responsible?
We are. Claude Code helps our engineers write and review code far faster, but architecture, business logic, security, testing and data quality are reviewed and owned by a person. Nothing reaches your systems because an AI produced it and nobody looked.