Module 10 Wrap-up: The Agent Architect

You have moved beyond the "Input/Output" model of AI. You understand that Agents are autonomous, goal-oriented brains that can reason through multi-step problems using Logic and Tools. You have learned how to structure their mission using clear, verb-driven instructions.

---

Hands-on Exercise: The Agent Blueprint

1. The Scenario

You are building an AI agent for a real estate agency.

• The Users: Home buyers.
• The Data: Available house listings.
• The Task: Help users find a home that fits their budget and location.

2. The Task

Write the Agent Instruction for this bot. Include:

1. A professional Persona.
2. A clear Success Criteria (e.g., "The task is complete when the user has selected a property to tour").
3. At least two Constraints (e.g., "Do not show houses above $1M").

---

Module 10 Summary

• Agents: Autonomous executors of workflows.
• Architecture: Instructions + Action Groups + Knowledge Bases.
• Reasoning Loop: How Bedrock decides which tool to call next.
• Instructions: The "Rules of Engagement" for the agent.

Coming Up Next...

In Module 11, we give the agent "Hands." We will learn about Action Groups and how to connect our agent to AWS Lambda functions so it can actually interact with databases and external APIs.

---

Module 10 Checklist

• I can describe the difference between a Chatbot and an Agent.
• I have written a mission statement (Instruction) for an AI agent.
• I understand the 3 core parts of the Bedrock Agent architecture.
• I know why "Active Verbs" improve agent performance.
• I have identified a real-world task that could be automated by an agent.