Skip to content

Navigation and Today's Command

The bottom navigation

Every page after sign-in has the same bottom bar:

EntryOpensAlso lit while you are on
ReconcileReconcile
FetchFetch Centerschedules, connections, the connect wizard, CCTV, document extraction, the orchestrator and surveillance
SimulateSimulate
TwinsActionable Twinsthe remediation screens, decision artifacts and Today’s Command
AdminAdminShown only to an admin of the selected instance. Other people see their role as a badge in the same place.
UserUser Profilebilling

The bar is keyboard-operable. Each entry is a button that screen readers announce with its label and its selected state.

Right after you choose an instance, the cockpit starts a session fetch in the background. This is the first reading of your data. When it completes, the compliance, finance and operations twins run on its findings and the rules are read again. That is why Twins and Reconcile fill in without you starting anything. The session fetch runs once per session, whichever page you open first.

The FACE AI assistant

The round button with the Runink mark, at the bottom right of every page, opens the assistant. Its tooltip is FACE AI, and screen readers announce it as Open FACE AI assistant.

The panel suggests prompts that fit the page you are on. For example, on Fetch or Connections it offers Create new connection and Find extra docs. On any other page it offers Help me with this screen and What can I do here?. Your prompt is sent together with the name of the page you are on.

ControlLabel
Text boxAsk FACE…
SendSend message
MicrophoneStart Recording / Stop Recording. Your speech is submitted when you stop.
CameraTake a photo
AttachmentAttach a file

Your question appears as You asked: {question}, followed by the assistant’s rationale and recommendation. FACE AI could not answer means the request failed, and the reason follows it. No answer came back means the assistant finished but returned nothing to show.

If voice input is blocked, the panel says Voice input is unavailable in this browser or the microphone was blocked. Type your question instead.

Today’s Command

Command View (subtitle Daily Ops Briefing) is a one-page briefing. Open it from the Twins page: select Today’s Command, the last chip under ACTIONABLE TWIN STATUS. The back arrow (Go back) returns you to Twins.

The headline card offers two buttons:

  • Run Fetch runs a trace that ranks business risks and opportunities. If the server refuses the run, its reason is shown in the card.
  • View Last Trace opens Fetch.

The AI Command bar takes a typed or spoken instruction (Ask or command…, with Voice input and Send). Read aloud speaks the answer. The bar also has two shortcuts: Video opens Surveillance, and Voice chat opens a voice call.

The four sections below the bar each link to the stage they summarise:

SectionShowsLink
1. SYSTEM POSTURE (Health & Context)Each data domain’s grade, alerts and maturity.View Data Posture Center →
2. BUSINESS LOGIC & RISKS (Rules Recon)Ranked risks and opportunities, once a fetch has run.View Rules Recon →
3. ACTIVE HYPOTHESES (Simulation)Ranked risks and opportunities, once a fetch has run.View Simulate →
4. DECISION TWINS (Actions)The decisions you approved in Twins, grouped by category.View Twins →

Empty sections explain themselves:

  • Not assessed means no grades exist yet.
  • Nothing is connected yet means no data source is attached. Connect one from Connections.
  • No decisions yet means you should run a fetch to have the twins analyse your data.
  • Nothing approved yet means you have not approved anything in Twins yet.

What evidence did you use? opens a drawer. It shows the trace id, how many inferred domains and data objects were analysed, and the sources. Before any analysis has run, it says No analysis has run in this session yet, so there is nothing to evidence.

Behind this view: FetchService.RunFetch, PostureService.GetHealthScore, TwinsService.GetActionQueue and ModelService. See the API reference.