Propel AI Bootcamp
Course 2 - Module 2

Agentic Development in Practice

Orchestrate custom models, RAG, and MCPs in real workflows

Learning Objectives

  • Orchestrate custom models, RAG, and MCPs
  • Handle errors in multi-agent systems
  • Manage state across workflow steps

Overview

Apply all Course 1 components in real agentic workflows.

Topics Covered

Integration

  • Using custom models + RAG + MCPs in pipelines
  • GitHub Actions integration

Agent Coordination

  • Communication patterns
  • State sharing

Error Handling

  • Multi-agent error strategies
  • NeMo Guardrails

Control Flow

  • Loops and conditionals
  • Branching strategies
  • State management

Resources

Check Your Understanding

Question 1 of 1

What is the purpose of guardrails in agentic systems?