AI agents and agentic AI

An AI agent is a language model given a goal, a set of tools such as APIs, databases or a browser, and the ability to decide which step to take next. Instead of answering a single question, it works through a task: look something up, take an action, check the result and continue.

1. 1

   Goal and tools

   The agent receives a task and a defined set of actions it is allowed to take.

2. 2

   Plan and act

   It reasons about the next step, calls a tool, reads the result and decides what to do next.

3. 3

   Boundaries

   Permissions, spending limits and approval steps keep it inside safe limits, with every action logged.

• Handling routine operations such as rescheduling bookings
• Researching and compiling information across systems
• Screening and shortlisting in hiring workflows
• Back office tasks that span several tools

We scope agents narrowly, give them only the tools a task needs, and design approval points for anything irreversible. Reliability comes from good tools and clear limits, not clever prompts.