What Duplex can do, in each scope.
A new agent knows straight away what Duplex will do for it in that work area.
Status and explanations. It runs Read-only, at session start.
A first day with no wrong turns.
An agent that finds out by trying spends its first day being refused and calling your AI for work Duplex already knows. One that starts with the list goes straight to the right request.
A visitor asks whether the museum has step-free access.
As the area changes, every new session sees the current list.
What Duplex can do in a work area changes as your team switches things on and Duplex learns more. Each new session starts from the list as it stands that day.
In March, the visitor desk opens: Duplex answers what it knows and briefs your AI on the rest.
What an agent sees fits every boundary around it.
The list a new agent sees respects your rules, the desk’s own settings and the agent’s scope, so it is never offered more than it may use.
Your rules apply first, to every agent.
Follows your settings.
Each work area has its own settings for speed, depth, cost and confidence, and a model’s output can never change them.
- How a work area balances speed and costEach work area shows the balance it strikes between speed, depth and cost.
- Your team tunes speed, depth and costA quick triage desk and a careful legal review can each run at their own pace and price.
- How sure it must be to answer aloneYour team can see how sure Duplex must be before it answers without your AI.
- Your team sets the confidence barHigh-stakes work can demand more certainty before Duplex answers on its own.
- What Duplex can do, in each scopeA new agent knows straight away what Duplex will do for it in that work area.
See it on your own work.
From your own assistant to every agent your company runs, a pilot puts Duplex on your own work, with the measures agreed before it begins.