You own the accounts
We build and run systems in accounts registered to your business: your cloud, your automation platform, your AI provider account.
Everything about how we handle your information, systems and access, in one place and in plain terms. If something you need is not here, ask, and we answer in writing.
These hold for every system we build, whatever the size of the project.
We build and run systems in accounts your business owns. Billing, data and control stay with you.
Separate, revocable credentials with only the permissions each task needs, removed when the work ends.
Sending, deleting or changing key records waits for a named person, unless you ask us in writing to let an agent post routine items itself. Payments and bank details always wait for a person.
Inputs, outputs and actions are logged, so you can see what happened and why.
We review the AI provider's current business terms with you before any of your data is processed.
We sign them before any of your information is shared, if you want them.
Where possible, every part of the system lives in accounts your business owns. The one thing that usually leaves is the text a model needs to read, and only the part the task requires.
We build and run systems in accounts registered to your business: your cloud, your automation platform, your AI provider account.
Each system gets only the permissions its task needs: read-only where reading is enough, limited to specific folders, mailboxes or records.
Only the information a task needs is sent to an AI model. Where practical, personal or sensitive details are removed or masked first.
Actions with real consequences, like sending to customers, moving money or changing important records, can require a person's approval.
Inputs, outputs and actions recorded at a level agreed with you, while keeping logs free of unnecessary sensitive data.
We test systems against real examples, including awkward and adversarial ones, before they're used for real work.
Your data stays in your systems. An agent reads what a task needs, when it needs it, through the same kind of access a person would use.
In accounts your business owns: your cloud account, your automation platform, your workspace. The billing is in your name, so there is nothing to migrate if you change suppliers and nothing that stops working if we stop working with you.
The security architecture above shows what lives in your environment, what leaves it, and the controls on each step.
You choose the provider, with a recommendation for your case. These are the public pages where each major provider describes how it handles business data.
| Provider | Where its data terms are published |
|---|---|
| OpenAI (API and business plans) | Enterprise privacy |
| Anthropic (API and business plans) | Commercial terms · Privacy center |
| Google (Gemini on Google Cloud) | Data retention and governance |
| Microsoft (Azure OpenAI models in Microsoft Foundry) | Data, privacy and security |
| Amazon Web Services (Amazon Bedrock) | Data protection |
| Open-weight models on your own infrastructure | No third-party model provider: the model runs in your environment |
Provider terms change. We review the current terms of the provider chosen for your project with you before any of your data is processed. These links are for reference and are not a summary of those terms.
No design removes these risks completely. The aim is to reduce them to a level you're comfortable with, and to make problems visible when they happen.
| Risk | What it looks like | How designs respond |
|---|---|---|
| Wrong or invented answers | Confident output that is incorrect | Answers grounded in your documents with citations, outputs checked against your data, human review where accuracy matters |
| Manipulated inputs | Text in an email or document trying to steer the system (“prompt injection”) | Tight limits on what the system may do, instructions kept separate from content, approval for sensitive actions |
| Data leaking between users | An assistant revealing something a person couldn't otherwise see | Your existing permissions respected rather than pooling everything together |
| Behavior changing over time | Model updates or new kinds of input changing results | A saved test set re-run after changes, and monitoring of real use |
If you work under rules such as GDPR, HIPAA, financial services regulation or professional confidentiality duties, those requirements shape the design from the first conversation: which data may be used, which providers are acceptable, what must be logged and what needs human review.
We are not a law firm and don't give legal advice. For regulated data, your legal or compliance advisers should confirm the approach meets your obligations, and the design documentation is written to make that review straightforward.
Questions about a specific project? Email info@georgyconsulting.ai.
AI systems can make mistakes, and any system can fail. What matters is how fast we catch it and how we handle it.
We pause the agent or revoke its credentials. We build every system so it can be stopped without stopping your business.
We tell you promptly and in writing: what happened, what was affected and what we are doing about it.
The audit trail shows the inputs, outputs and actions involved, so we find the cause rather than guess it.
We test the fix on the case that failed and the existing test set before the system goes back into use, and add the case to the test set.
Book a free 20-minute call. You'll leave knowing exactly where AI pays off in your business, and what to do first.