What it is
Kroy is not only a gateway to other systems. It is a database in its own right. KroyDB stores organisational information as entities — a client, a project, a supplier, a policy, a decision — each with a schema, relationships to other entities and a permission model.
Typical entities include clients, projects, matters, products, services, contacts, suppliers, articles, policies, decisions, tasks, requests, assets, locations, FAQs, Skills and Artifacts.
Why it exists
A great deal of what an organisation knows has no proper home. It lives in spreadsheets, in email threads and in people’s heads. When an AI needs it, someone pastes it into a chat — and the next conversation starts from nothing.
KroyDB gives that information a structured, governed home that every authorised AI can reach.
How it works
- Import, don’t upload. A CSV client list becomes structured client entities, not a file sitting in a folder.
- Relationships are first-class. A client relates to its projects, contacts, deadlines and documents, so an AI can answer questions that span them.
- Permissions travel with the data. Every entity is subject to Kroy’s policies, so what a person or agent can see is decided by Kroy, not by what happened to be pasted.
- Store or connect. Where a system such as Xero remains authoritative, Kroy connects to it instead of copying it.
What permissions apply
Access to every entity is resolved by Kroy Control: the person, the agent acting for them, the delegation between them and the organisation’s policies. Different audiences receive different projections of the same entity.