Why roothr.ai is built on Root HR, not beside it

There is a tempting way to build an HR assistant: connect it to everything, let it read spreadsheets and email and whatever HR system a customer happens to have, and hope the model sorts it out. We chose the opposite. roothr.ai reads one system, Root HR, and it does so through the same permissions a person would have.

One source of truth

When you ask who is on leave on Friday, there is exactly one place that answer lives. That means the assistant does not have to reconcile three versions of a rota, and it means you can check its answer in the same place it looked.

Permissions are not optional

A line manager asking about pay should see their team, not the company. That is already how Root HR works, and roothr.ai inherits it rather than reimplementing it. The assistant runs as the person asking, so it can never reveal more than they could open themselves.

Every answer shows its working

Figures link to the records they came from. A compliance flag names the worker, the document and the date. If the assistant is not sure, it says so and asks rather than filling the gap with a plausible guess. In payroll and immigration, a confident wrong answer is worse than no answer.

Same infrastructure, same controls

roothr.ai runs on the same UK-hosted infrastructure as Root HR, under the same ISO 27001 controls, ICO registration and GDPR arrangements. Prompts and answers are processed within that boundary and are not used to train third-party models.

None of this is glamorous. It is the difference between an assistant you can use for a demo and one you can hand your payroll to.