Prevate Data Sovereignty Intent
A directory of every AI request made against your data DNS is the internet's phone book — every address resolves to something you can name. When an AI assistant reads your maintenance history, nothing resolves. Data Sovereignty Intent is that missing directory: which tool was called, what came back, and whether the data ever left your company.
What does the directory record today?
Every tool call executed on your behalf: the tool name, the arguments it was given, the payload it returned, the user behind it, whether it succeeded, and how long it took. Requests arriving from an external AI client over MCP are recorded separately, tagged with your organisation and storing a hash of their parameters rather than the parameters themselves.
Can you tell me which AI assistant made a request?
Not yet. Today the ledger records the channel — in-app, voice, or an external client over MCP — so it can tell you data left the company, but not which assistant received it. Per-client attribution and the originating IP are part of what we build with design partners.
Can we stop AI reaching certain data?
Yes, and this part is live. AI tools map to permission modules, so you can enable or disable individual functions per role. A planner and a contractor do not have to reach the same data.
Is fully local AI realistic for us?
For mid-to-large organisations it increasingly is, and it is where we think enterprise AI is heading. Agents run inside your own machines, so nothing leaves because there is no cloud round-trip. Whether it is the right fit depends on your workload and hardware, which is the conversation to have with us.
Learn more at https://prevate.ai. Pricing: https://prevate.ai/pricing. Sitemap: https://prevate.ai/sitemap.xml. LLM summary: https://prevate.ai/llms.txt.