Agents

Provider

From “Ready for Discharge” to Safely Home

Coordinate discharge readiness, patient education, medication management, follow-up scheduling, and bed release to create a faster and safer transition from hospital to home.

Outcome

A faster, coordinated and safer transition from hospital to home—while improving patient experience and making beds available sooner.

Agents at work
CarerVoiz· 2 active
CarerKMS· 3 active
Hospilot· 2 active

Tap a node for its name

Under the hood

How Carer Agents Think, Collaborate & Act

Carer agents don't work in isolation — they discover, evaluate, collaborate and coordinate dynamically to achieve a shared goal. Built on an agentic orchestration layer, they move intelligently from intent to decision to governed action.

01Discover

Dynamic Agent Discovery

Agents publish their capabilities, tools, context requirements and permissions. The orchestration layer dynamically discovers the right agents for each task.

Event→Discover→Match
02Bid

Agent Bidding & Selection

Multiple agents evaluate the same task and “bid” based on specialization, context, confidence and availability. The orchestrator selects the most appropriate agent — or team.

Task→Bid→Evaluate→Select
03Compose

Dynamic Workflow Creation

Workflows don’t need to be predefined. Based on the healthcare event, available agents and policies, Carer composes an execution path on the fly.

Goal→Compose→Build→Execute
04Delegate

Agent-to-Agent Collaboration

Agents delegate subtasks, exchange structured context and request capabilities from one another while maintaining a common workflow state.

Delegate→Collaborate→Resolve
05Collaborate

Multi-Agent Orchestration

Specialized agents collaborate rather than operate in isolation. An orchestrator coordinates dependencies, execution order, shared context and outcomes across agents.

Plan→Coordinate→Execute
06Remember

Context & Shared Memory

Agents maintain context across conversations, decisions, workflow events and enterprise systems — so one agent can continue exactly where another left off.

Context→Memory→Continuity
07Approve

Human-in-the-Loop Governance

High-impact actions can require human review. Agents run in Assisted, Advisory or Autonomous modes depending on workflow, confidence, permissions and policy.

Recommend→Approve→Act
08Act

Policy-Governed Autonomous Actions

Every agent operates within defined permissions, policies and action boundaries. Approved agents execute across healthcare systems while keeping a full audit trail.

Policy→Permission→Action→Audit