Planner
A planner is the component of an AI agent that turns a goal into an ordered sequence of steps before any action is taken, deciding what to do in which order instead of improvising one move at a time.
No jargon, no math
In plain words, what is it?
Acting without a plan means the agent picks its next move from whatever it just saw, which works for two steps and falls apart by five. A planner works like a good site manager: before anyone lifts a panel, the job is broken into ordered tasks with dependencies, tools assigned, and checkpoints where a human must sign off. It reads the goal, lists everything that needs to happen, orders it (you cannot invoice before you have the delivery note) and estimates what each step needs. Then the executor works the plan. Where it breaks: the plan is only as good as the planner's picture of the world. Miss a constraint, like the lab being closed on Friday, and a beautiful plan walks straight into a wall. That is why plans get re-checked against reality as steps complete.
Terms that live nearby.
Agentic Workflow
An agentic workflow is a business process whose steps are executed by AI agents: they plan the work, call tools, read results and decide the next move, instead of following a fixed script that breaks the first time reality differs.
ReadOrchestration
Orchestration is the coordination layer of a multi-agent system: one orchestrator receives a goal, splits it into tasks, assigns them to specialised agents, collects their outputs, resolves conflicts and assembles the final result.
ReadAgent Loop
The agent loop is the cycle an AI agent repeats while pursuing a goal: observe the current state, plan the next action, act through a tool, check the result, then continue, re-plan or stop.
Read§ 03
Want this running in your stack?
We design and ship AI systems with permissions, logs and human approval. It starts with a 30-minute scoping call.