Your Data Stays Under Your Control.
Deploying AI inside a business means working with real data. Here's how we handle it: agreed access, minimum permissions, people approving what matters, and a record of everything the AI does.
Security Is Part of the Design
We decide what data a workflow can reach, where it goes and who approves the output before we build, not after. That's what lets your IT team say yes with confidence, and it's why data questions are part of every AI Workflow Assessment.
Our Commitments
The principles we apply to every AI implementation project.
- Data access agreed in writing before production data is used
- Only the data a workflow needs, nothing more
- Credentials your team controls and can revoke
- Human approval for anything customer-facing
- Every AI action logged and reviewable
- Your data is never used for our marketing
How Your Data Moves Through a Project
- 1
Scoping
During the assessment we identify which systems and data a workflow needs, and record it.
- 2
Agreement
Access, processing location, AI model providers and retention are set out in the project agreement.
- 3
Access Set Up
Integrations use dedicated, least-privilege credentials owned by your IT team.
- 4
Build and Test
Development uses sample or masked data wherever possible before any production data is connected.
- 5
Run With Oversight
Live workflows keep approval steps, escalation rules and a full action log.
- 6
Return or Delete
At the end of an engagement, data we hold is returned or deleted as agreed.
Controls We Build In
Access Control
Role-based access and least-privilege integration accounts.
Encryption
Data encrypted in transit and at rest.
Approval Steps
People sign off on outputs that matter.
Audit Logs
What the AI read, proposed and changed.
Model Provider Review
Which AI providers are used, and on what data terms, agreed with you in advance.
Monitoring
Failures and unusual behaviour surfaced to a person, not hidden.
Regulations We Design Around
We build to the requirements that apply to your business and your customers.
Questions Your IT Team Can Ask Us
We’re happy to answer these, in writing, before any project starts.
- Which of our systems will the AI access?
- Where will our data be processed and stored?
- Which AI model providers are involved?
- Can those providers retain or train on our data?
- How do we revoke access?
- What happens to our data when the project ends?
Have a Security Review to Complete?
Send us your questionnaire or your IT team's questions. We'll answer them in writing before any project begins.