Agentic AI
by DeepLearning.AI
Andrew Ng's hands-on course on building agentic workflows with the four core design patterns.
Overview
Agentic AI, taught by Andrew Ng on the DeepLearning.AI platform, teaches developers to build LLM-powered agents that complete complex, multi-step tasks through iterative workflows instead of one-shot prompting. The course is organized around four agentic design patterns—Reflection (an agent critiques and revises its own output to improve quality), Tool Use (connecting the model to databases, APIs, code execution, and external services), Planning (decomposing a goal into an executable sequence of steps), and Multi-Agent collaboration (coordinating specialized agents on subtasks). Across roughly 10 hours, 31 video lessons, 7 code examples, and 8 graded assignments, it moves from evaluating and debugging agentic systems to composing these patterns into production-style applications. It targets intermediate developers who know Python and have basic familiarity with LLM APIs, and is available with a DeepLearning.AI Pro membership with a certificate on completion.
At a Glance
- Topic
- Agentic
- Level
- Intermediate
- Format
- Course
- Cost
- Paid
- Duration
- ~10 hours, self-paced (31 lessons across 5 modules)
- Provider
- DeepLearning.AI
- Hands-on
- Yes — code/exercises
- Certificate
- Available
What You’ll Learn
- ✓Apply the Reflection pattern so an agent evaluates and improves its own output
- ✓Give agents Tool Use to call APIs, databases, code execution, and external services
- ✓Use Planning to decompose complex goals into executable multi-step workflows
- ✓Coordinate Multi-Agent systems where specialized agents collaborate on subtasks
- ✓Evaluate, debug, and iterate on agentic workflows to improve reliability
Highlights
- •Taught by Andrew Ng, founder of DeepLearning.AI
- •Framework-agnostic treatment of the four canonical agentic design patterns
- •Practical: 7 code examples and 8 graded, hands-on assignments
- •Emphasis on evaluation and iteration, not just building
Who It’s For
Best For
- ✓Developers moving from single-prompt apps to multi-step agentic systems
- ✓AI engineers who want a rigorous, pattern-based mental model for agents
- ✓Python developers with basic LLM API experience
Prerequisites
- •Intermediate Python
- •Basic familiarity with LLM APIs and prompting
FAQ
What is Agentic AI?
A hands-on course by Andrew Ng for developers who want to build agentic AI systems that plan, use tools, and iterate rather than emit single responses. It teaches the four foundational agentic design patterns with Python code labs and graded assignments.
Is Agentic AI free?
Agentic AI is a paid resource.
What level is Agentic AI for?
Agentic AI is aimed at a intermediate audience. Recommended background: Intermediate Python, Basic familiarity with LLM APIs and prompting.
How long does Agentic AI take?
Expect roughly ~10 hours, self-paced (31 lessons across 5 modules). Most learners work through it at their own pace.
What will I learn from Agentic AI?
You'll learn: Apply the Reflection pattern so an agent evaluates and improves its own output; Give agents Tool Use to call APIs, databases, code execution, and external services; Use Planning to decompose complex goals into executable multi-step workflows; Coordinate Multi-Agent systems where specialized agents collaborate on subtasks; Evaluate, debug, and iterate on agentic workflows to improve reliability.