Thinking / AI agents
What is an AI agent, and does your business actually need one?
A clear guide to where autonomous action adds value and where a simpler assistant, automation or process change will work better.
The answer in brief
- What it is
- A clear guide to where autonomous action adds value and where a simpler assistant, automation or process change will work better.
- Best suited to
- Leaders who need to translate this issue into an investment, workflow, governance or capability decision.
- What useful progress looks like
- Products described as agents range from conversational assistants to systems that can retrieve information, plan work, call tools and take action across applications. That ambiguity encourages organisations to select a technical pattern before describing the work. The practical distinction is agency. An assistant produces material for a person to use. A deterministic automation follows predefined rules. An agent interprets context and selects actions within a delegated boundary. Greater agency can create value across variable work, while it also increases the need for explicit objectives, permissions, monitoring and recovery.
The question leaders ask
Does our business need an AI agent?
Use an agent when valuable work requires a system to interpret changing context, choose among permitted actions and complete several steps under clear guardrails. Use simpler automation when rules and paths are stable.
01
Agent has become an overloaded label
Products described as agents range from conversational assistants to systems that can retrieve information, plan work, call tools and take action across applications. That ambiguity encourages organisations to select a technical pattern before describing the work. The practical distinction is agency. An assistant produces material for a person to use. A deterministic automation follows predefined rules. An agent interprets context and selects actions within a delegated boundary. Greater agency can create value across variable work, while it also increases the need for explicit objectives, permissions, monitoring and recovery.
02
Look for three characteristics
A credible agent opportunity usually combines context, choice and action. Remove one of those elements and a simpler pattern may offer better reliability and cost.
Context
The system must assemble relevant information from documents, systems or a live interaction.
Choice
The next step depends on interpretation rather than a fixed sequence of rules.
Action
The system must update a record, communicate, route work, trigger a tool or complete another permitted step.
Repetition
The workflow occurs often enough for improved speed or consistency to matter.
Boundaries
Permitted actions, prohibited actions and escalation conditions can be expressed clearly.
Evidence
Quality, time, customer or commercial outcomes can be measured against a baseline.
03
Choose the lightest capable solution
Some problems improve when information becomes easier to find. Others need a structured form, a clearer approval or a conventional workflow automation. An agent introduces model variability, continuing evaluation and a larger security surface. Leaders should compare process change, existing platform capability, deterministic automation, an AI assistant and an autonomous agent. The selected approach should be the least complex option able to create the intended outcome under acceptable control.
04
Design the operating system around it
Production value depends on more than a strong demonstration. The business must define what the agent optimises, which information it may access, which actions it can take, when it asks a human and how an incorrect action is reversed. Someone must own performance after launch. Users need to understand their authority and responsibilities. Monitoring should cover business outcomes, decision quality, incidents, cost and changing behaviour over time. These conditions turn an impressive prototype into a dependable part of work.
FAQ
Questions leaders ask.
Direct answers to the questions that commonly shape an initial conversation.
01Is a chatbot an AI agent?+
A chatbot may only answer questions. It becomes more agentic when it selects and completes actions across systems within delegated boundaries.
02What is a good first agent workflow?+
Choose repeatable work with a clear owner, accessible context, measurable value, limited permissions and an obvious human escalation route.
03Can we build an agent with existing software?+
Often yes. CRM, service, marketing and productivity platforms increasingly include agent capabilities. Their fit, data access, controls and continuing cost still require assessment.
04How long should an agent proof take?+
A focused proof can often be designed and tested within four to eight weeks when the workflow and information are ready. Production hardening may require longer.
Latest from the blog
Useful ideas for the decision in front of you.
Plate Spinner's PerspectiveFF＝01
Plate Spinner's Perspective · 8 minLearning to disagree with AI
Why a confident, persuasive answer should be the start of leadership judgement, not the end of it.
Read the articlePlate Spinner's PerspectiveFF＝02
Plate Spinner's Perspective · 7 minWhat does a Chief of Staff do in an AI business?
The operating role between technical possibility, commercial pressure and executive attention.
Read the articlePlate Spinner's PerspectiveFF＝03
Plate Spinner's Perspective · 7 minChief of Staff vs COO: where does the work split?
A practical distinction between enterprise operations and the executive agenda that cuts across them.
Read the articleApply the thinking
Find the right first workflow for an agent
Bring the live decision, workflow or commercial pressure. We will help translate the idea into a focused next step.
Discuss the implication