AI
No back door, by construction.
The assistant does not have its own view of the data. It queries as you, through the same permission checks as every screen, which is why there is no separate AI permission model to keep in step.
One permission model
If a second model existed it would eventually diverge from the first, and the divergence would be a data breach. So there is only one.
Sensitive categories are opt-in
Compensation, health-related absence and private manager notes are excluded from assistant responses by default even for people who could open them directly.
Questions are recorded where it matters
Asking about sensitive categories is logged like viewing them. The convenience of asking should not be a way around the audit trail.
AI
The rest of this area.
Questions in normal language, answered from the record, with the source shown.
Workforce assistant
Ask about the business, not about the software.
Daily briefing
What needs attention across the team today.
What it can answer
The questions this structure makes answerable.
Where answers come from
Every answer cites the record behind it.
Suggestions
What Orglume noticed, why it matters, and what you could do.
What it will not do
The line between assistance and employment decisions.
Run your team with clarity.
From the moment someone joins to the day their responsibilities are handed over, Orglume keeps your people, their work and your organization connected.
Set up from a spreadsheet · Nothing about the AI is metered · [email protected]