Security
Your access rules apply before the AI sees anything. It runs in your region, on your own servers or in your cloud.
Our founding team has more than 30 years of experience protecting sensitive data in healthcare and banking. This page sets out how Platform handles yours: who can see what, where it runs, and what is recorded.
Your access rules, applied first.
Who can see what comes from your own systems, the same access your teams already manage. Those rules apply before the AI sees anything. Information a person can't open is never sent to a model on their behalf, and it doesn't appear in their answers or counts. If a source's access information stops updating, that source stops answering rather than risk showing too much.
Read-only.
Platform reads your systems and never changes them. It doesn't send, approve or edit anything in another system on anyone's behalf.
You control who connects.
It uses your own certificates for encrypted connections, and only the servers you approve can connect to it. Your IT team sets both.
A record of who saw what.
Documents open only when they're needed, and every opening is recorded: who saw what, when, and for which question.
Runs where your data lives.
Platform runs in your region, on your own servers or in your cloud, so your data stays under your control. Each organisation's data is kept separate. Because it runs in your environment, it sits within your existing security controls and certifications.
Your choice of AI model.
You choose the AI model and provider, and it works under the same rules. We can recommend options for your region, your data, safety and cost. Only the information needed for a question, and allowed for that person, is sent to it.
Only what your records show
AI models can produce answers that sound right but aren't. Platform only states what a record supports. Each statement links to the record behind it, so anyone can check it, and if the records don't answer the question, it says so instead of guessing.
Talk through security with us.
We'll walk your security team through the details and answer your security questionnaire.
Talk to us