Built to pass your security, audit and data protection reviews
Controls are designed in from the first sprint and documented for your information security and data protection teams.
Data residency
- Deployed in your tenant and approved cloud region
- No client data copied into our systems
- AI models accessed through your enterprise endpoints
Information security
- Aligned with ISO/IEC 27001 control objectives
- Single sign-on and MFA through your identity platform
- Least-privilege, role-based access; encryption in transit and at rest
Responsible AI
- Practices aligned with the NIST AI Risk Management Framework and ISO/IEC 42001
- Answers grounded in approved sources, with citations
- Each agent tested against an evaluation set before release
Human oversight
- Actions above a defined risk threshold wait for approval
- Every automated action logged: who, what, when, source
- Kill switch and rollback for every agent
Personnel
- Named team members only, each under NDA
- Access granted per engagement and revoked at exit
- Background checks supported on request
Data protection
- Designed to comply with applicable national data protection law
- Data protection impact assessment support
- Retention and deletion rules configured per policy
Assurance documentation pack
Available under NDA before you commit, and delivered with every release.
What clients usually ask us
Can you work alongside our IT team and existing vendors?
Yes. We contract directly or as a specialist partner within your existing vendor arrangements, and we work inside your change and security processes.
Where is our data stored and processed?
In your own tenant and approved cloud region. We do not copy client data into our systems, and AI models are accessed through your approved enterprise endpoints.
Who owns what you build?
You do. All code, configuration, data models, prompts and documentation transfer to you with each release.
How do you control AI errors?
Answers are grounded in your approved sources and show citations. Every agent is tested against an evaluation set before release, and actions above a defined risk threshold require human approval.
Can the assistants run on our collaboration tools?
Yes. Agents run on Microsoft Teams, Outlook, WhatsApp Business or your approved internal channel. The channel is chosen to match your security policy.
What happens if we end the engagement?
You keep everything. The exit plan includes runbooks, architecture documentation and a handover period so your team or another supplier can operate the system.
What does a pilot cost?
Pilots are fixed-price and scoped after the assessment, so you know the cost, timeline and success measure before work starts.
See these systems working in a 45-minute demo
Live demonstrations, a discussion of your priorities, and a written proposal for an assessment or pilot within five working days.