Find your next step.

Ctrl / ⌘ K to open · Esc to close · Demonstration commands use fictional data.

OPERATE / OBSERVABILITY

Every agent.
One operational view.

Operational visibility should connect agent health, task outcomes, model usage, latency, costs, incidents, policy blocks, and human escalations. See where work is moving and where it needs attention.

Explore the live demoPlatform architecture
INTELLIGENCEOS / OPERATE01 — OVERVIEW

02 / THE OPERATING CONTRACT

What observability
brings into view.

01

Task health

Separate running, waiting, blocked, failed, and completed tasks. A task waiting on approval should not be reported as a technical failure.

02

Model and tool usage

Inspect which capabilities and systems a workflow used. Capture latency and tool errors without hiding the business context.

03

Cost intelligence

Track compute, model, tool, and task costs with explicit attribution. Optimization suggestions need quality and privacy checks.

04

Incident timelines

Connect a failure to its agent version, tool request, policy decision, and recovery step. Route unresolved events to a responsible owner.

03 / CONNECTED BY DESIGN

The path
through the system.

Inspect the sequence.
Each stage has a defined responsibility.

Select a node to inspect its role in the workflow.

Before work begins.

Define the source of truth, permitted actions, accountable owner, and acceptance criteria. Scope matters more than an open-ended instruction.

Before work moves on.

Check the outcome, capture unresolved exceptions, and preserve the relevant policy and evidence. A task is complete only within its defined boundary.

04 / A CONCRETE EXAMPLE

A healthy workforce can still have blocked actions.

Northstar’s illustrative dashboard separates policy blocks from tool failures and review queues. A prohibited request stopped by FIN-04 is a functioning boundary. A data source timeout is a reliability incident. Each has a different owner and next step.

ILLUSTRATIVE SCENARIO

Use verified business outcomes alongside latency and cost. A fast, cheap task that used the wrong data is not a successful task.

Defined scopeEvidence preservedReview when required
Try the control plane

05 / QUESTIONS & BOUNDARIES

Details
that matter.

Architecture descriptions show intended system behavior. Validate implementation, integration, data handling, and operational controls before relying on a production workflow.

Are dashboard percentages customer results?

No. Any displayed agent counts, task totals, and percentages are fictional interface examples.

What should an incident contain?

The affected task, relevant versions, timing, data and tool context, impact, owner, and the recorded recovery action.

How does this fit with the rest of IntelligenceOS?

Identity, context, policy, evaluation, verification, execution, and audit form one operating model. Explore the related components below to see their responsibilities.

06 / CONTINUE THE EXPLORATION

See the whole
operating picture.