Agentic AI

Prompting for Effective LLM Reasoning and Planning

• Lesson 1: Introduction to Prompting for Effective LLM Reasoning and Planning
  Introduces the core concepts of Agentic AI, the course structure, prerequisites and learning environment.
• Lesson 2: The Role of Prompting in Agentic AI with Python and OpenAI
  Learn what AI agents are and how prompting guides them to reason, plan and act toward goals.
• Lesson 3: Role-Based Prompting
  Explore how roles and personas shape tone, expertise and behaviour in agent outputs.
• Lesson 4: Implementing Role-Based Prompting with Python
  Practise building role-based prompts to create a believable historical persona.
• Lesson 5: Chain-of-Thought and ReAct Prompting
  Understand frameworks for guided reasoning and action-oriented planning.
• Lesson 6: Applying CoT and ReAct Prompting with Python
  Implement reasoning and action prompts to solve a retail analytics problem.
• Lesson 7: Prompt Instruction Refinement
  Learn systematic approaches to refining prompts across role, task, context and output.
• Lesson 8: Applying Prompt Instruction Refinement with Python
  Iteratively refine a prompt to transform a generic tool into a structured dietary consultant.
• Lesson 9: Chaining Prompts for Agentic Reasoning
  Design multi-step workflows by chaining prompts with validation checkpoints.
• Lesson 10: Chaining Prompts with Python
  Build a three-stage prompt chain with gate checks to automate insurance claim triage.
• Lesson 11: LLM Feedback Loops
  Design systems where agents improve outputs through iterative feedback.
• Lesson 12: Implementing LLM Feedback Loops with Python
  Create a self-debugging agent that generates and tests Python code.

Project: AgentsVille Trip Planner: Build a multi-agent travel assistant that coordinates planning, reasoning and execution.

Introduces the core concepts of Agentic AI, the course structure, prerequisites and learning environment.

Agentic Workflows

• Lesson 1: Introduction to Agentic Workflows
  Explore the foundational concepts of agentic workflows and system setup.
• Lesson 2: Understanding Agentic Workflows
  Learn the core components of modern AI agents and workflow types.
• Lesson 3: Agentic Workflow Modelling
  Design and visualise agentic workflows using common agent patterns.
• Lesson 4: Agentic Workflow Implementation
  Translate workflow models into Python code and orchestrate agent interactions.
• Lesson 5: Prompt Chaining Workflow Pattern
  Break down complex tasks into sequential agent workflows.
• Lesson 6: Implementing Prompt Chaining with Python
  Build a multi-agent chain that passes information step by step.
• Lesson 7: Routing Workflow Pattern
  Design routing systems that dispatch tasks to specialised agents.
• Lesson 8: Implementing Routing with Python
  Create a routing agent that classifies requests and orchestrates responses.
• Lesson 9: Parallelisation Workflow Pattern
  Run multiple agent tasks concurrently and aggregate results.
• Lesson 10: Implementing Parallelisation with Python
  Use Python threading to build concurrent agent workflows.
• Lesson 11: Evaluator–Optimiser Workflow Pattern
  Iteratively improve outputs through critique and refinement loops.
• Lesson 12: Implementing Evaluator–Optimiser with Python
  Build a creator–critic system that refines outputs until constraints are met.
• Lesson 13: Orchestrator–Workers Workflow Pattern
  Design systems where a central agent plans and delegates work dynamically.
• Lesson 14: Implementing Orchestrator–Workers with Python
  Build an agentic market analysis system with specialised workers.

Project: AI-Powered Agentic Workflow for Project Management: Create a reusable agent library and deploy it to manage a technical project.

Building Agents

• Lesson 1: Introduction to Building Agents
  Set up tools and environments for agent development.
• Lesson 2: Extending Agents with Tools
  Integrate tools to enable real-time actions and data access.
• Lesson 3: Building Agents with Tools in Python
  Develop and test tool-enabled agents using the OpenAI SDK.
• Lesson 4: Structured Outputs
  Transform agent responses into validated, machine-readable outputs.
• Lesson 5: Implementing Structured Outputs with Pydantic
  Create validated JSON outputs for reliable automation.
• Lesson 6: Agent State Management
  Design state machines that track context and decisions across workflows.
• Lesson 7: Implementing Agent State with Python
  Build dynamic workflows using Python state machines.
• Lesson 8: Short-Term Agent Memory
  Implement memory strategies to maintain coherence in active sessions.
• Lesson 9: Adding Agent Memory with Python
  Build persona-based agents with session continuity.
• Lesson 10: External Tools and APIs
  Integrate external APIs for real-time data and actions.
• Lesson 11: Integrating External Tools with Python
  Build agents that authenticate and interact with live systems.
• Lesson 12: Web Search Agents
  Enable agents to retrieve and ground responses in real-time information.
• Lesson 13: Creating Web Search Agents with Python
  Build a web-enabled agent using external search APIs.
• Lesson 14: Interacting with Databases
  Connect agents to SQL and vector databases.
• Lesson 15: Building Database Agents with Python
  Translate natural language into database queries.
• Lesson 16: Agentic Retrieval Augmented Generation
  Enhance RAG systems with reflection and retry loops.
• Lesson 17: Agentic RAG with Python
  Build adaptive document retrieval systems.
• Lesson 18: Long-Term Agent Memory
  Design persistent memory strategies for agents.
• Lesson 19: Maintaining Long-Term Memory with Python
  Implement long-term memory using vector databases.
• Lesson 20: Agent Evaluation
  Assess agent performance, quality and reliability.
• Lesson 21: Evaluating Agents with Python
  Design evaluation frameworks and test cases.

Project: UdaPlay AI Research Agent: Build a stateful research agent for the video game industry.

Multi-Agent Systems

• Lesson 1: Introduction to Multi-Agent Systems
  Learn the fundamentals of multi-agent architectures.
• Lesson 2: Designing Multi-Agent Architecture
  Design high-level system architectures for agent teams.
• Lesson 3: Implementing Multi-Agent Architecture with Python
  Code multi-agent systems with well-defined interfaces.
• Lesson 4: Orchestrating Agent Activities
  Coordinate agent actions across complex workflows.
• Lesson 5: Implementing Agent Orchestration
  Apply orchestration patterns for multi-step execution.
• Lesson 6: Routing and Data Flow in Agentic Systems
  Manage information flow between agents.
• Lesson 7: Implementing Routing and Data Flow
  Build intelligent routing agents.
• Lesson 8: State Management in Multi-Agent Systems
  Track and update state across agent interactions.
• Lesson 9: Implementing State Management
  Synchronise state across multiple agents.
• Lesson 10: Multi-Agent Orchestration and Coordination
  Resolve conflicts and manage shared resources.
• Lesson 11: Implementing Orchestration and Coordination
  Build concurrent multi-agent systems.
• Lesson 12: Multi-Agent Retrieval Augmented Generation
  Extend RAG systems across agent teams.
• Lesson 13: Implementing Multi-Agent RAG
  Build cooperative retrieval and synthesis agents.

Project: The Beaver’s Choice Paper Company Sales Team: Design and implement a complete multi-agent sales system.