What you'll learn

What you'll learn:

• Orchestrator agents: Coordinate other agents
• Worker agents: Execute specific tasks
• Tool agents: Wrap external APIs
• Memory agents: Store and retrieve context

Minor reframing:"Roles ≠ Names

Roles describe what an agent does (orchestrate, work, store).Names describe identity (Shopping Agent, Payment Agent).

Same agent can have different roles:

• Shopping Agent might be a worker during discovery
• Shopping Agent might be an orchestrator during checkout

Or roles can be separate agents with handoffs:

• Orchestrator Agent (coordinates)
• Shopping Agent (worker)
• Payment Agent (worker)

This is the role-based vs handoff-based choice."

Types of Agents

Understanding roles inside a multi-agent system

Multi-agent systems only work when roles are clearly defined.
When roles blur, systems become unpredictable, unsafe, and impossible to audit.

This section explains the four most common agent roles used in modern AI products — and how each one affects UX, safety, and system stability.

Why Agent Roles Matter

Humans understand systems through roles:

• Who is responsible?

• Who decides?

• Who executes?

• Who remembers?

Multi-agent systems must reflect this logic.

Without explicit roles:

• multiple agents attempt the same task

• responsibilities overlap

• errors propagate

• recovery becomes impossible

• UX feels chaotic

Clear roles create:
Predictability
Accountability
Safe autonomy
Meaningful control points

1. Orchestrator Agent (The Coordinator)

What it is

The Orchestrator is the agent responsible for planning, coordination, and delegation.

It does not usually perform tasks itself.

What it does

• Interprets the user’s overall intent
• Breaks goals into sub-tasks
• Decides which agents to involve
• Manages sequencing and dependencies
• Handles retries, fallbacks, and escalation

What it should NOT do

• Call external tools directly (usually)
• Generate detailed content
• Own domain-specific logic

Why it matters for UX

The orchestrator determines:

• What happens first
• What happens next
• When the user is asked for confirmation
• When failures surface

Poor orchestration leads to:

• confusing flows
• silent loops
• missing steps
• unexpected actions

UX rule:
The user should feel guided, not shuffled.

2. Worker Agents (The Specialists)

What they are

Worker agents handle specific tasks within a constrained domain.

Each worker should be good at one thing.

Examples

• Writing agent
• Classification agent
• Recommendation agent
• Booking agent
• Risk-evaluation agent

What they do

• Perform focused reasoning
• Generate outputs within scope
• Apply domain rules
• Return results to the orchestrator

What they should NOT do

• Make high-level decisions
• Coordinate other agents
• Change task direction
• Act autonomously beyond their assignment

Why this matters for UX

Specialization improves quality only if boundaries are respected.

Failures occur when:

• workers improvise outside scope
• multiple workers overlap responsibilities
• outputs conflict

UX rule:
Users should never experience internal disagreements between agents.

3. Tool Agents (The Executors)

What they are

Tool agents are responsible for calling external systems.

They act as controlled bridges between AI reasoning and real-world actions.

Examples

• API callers

• Database query agents

• Payment processors

• Calendar updaters

• File handlers

What they do

• Execute actions

• Enforce constraints

• Return structured results

• Report errors clearly

What they should NOT do

• Reason about goals

• Interpret user intent

• Decide whether to act

Tool agents should be predictable and deterministic.

Why this matters for UX

Most serious failures come from improper tool use:

• accidental purchases

• data modification

• irreversible actions

Users do not blame “tool agents.”
They blame the product.

UX rule:
All tool actions must be intentional, transparent, and recoverable.

4. Memory Agents (The State Keepers)

What they are

Memory agents manage:

• context
• preferences
• task state
• session history

They enable continuity across time.

What they do

• Store and retrieve relevant information
• Decide what is worth remembering
• Expose memory safely to other agents

What they should NOT do

• Guess personal traits
• Store sensitive information without consent
• Apply memory invisibly

Why this matters for UX

Memory dramatically improves efficiency — but also increases risk.

Common failures:

• incorrect recall
• context drift
• over-personalization
• “creepy” assumptions

UX rule:
If memory affects behavior, users must understand and control it.

5. Optional / Advanced Agent Roles (Briefly)

You may also encounter:

• Safety agents (policy enforcement)

• Evaluation agents (quality checks)

• Monitoring agents (observability & logging)

• Escalation agents (handoff to humans)

These are typically backstage roles and should rarely surface to users directly.

Common Role Design Mistakes

Avoid these patterns:

• One agent doing everything

• Workers making autonomous decisions

• Tool agents acting without confirmation

• Memory applied without transparency

• Orchestrator exposed directly to users

These lead to:

• broken trust

• unpredictable behavior

• audit failure