Hands-on activities to reinforce your learning and build real agent components
Practical application is at the heart of this course. Each module includes hands-on exercises designed to reinforce theoretical concepts and develop your skills in building agentic AI systems.
These exercises range from guided tutorials to open-ended projects, gradually increasing in complexity as you progress through the course. By completing these exercises, you'll build a portfolio of agent components and eventually a complete agentic AI system.
All exercises include starter code, clear instructions, and evaluation criteria to help you assess your progress. Many exercises also offer extension challenges for those looking to deepen their understanding.
Analyze and compare different types of AI agents in real-world applications.
Map out the capabilities and limitations of different agent architectures.
Develop an ethical framework for agentic AI system development.
Implement a basic perception module for processing textual input.
Build a simple reasoning component using rule-based logic.
Integrate perception and cognition components into a simple agent.
Create a knowledge graph for a domain-specific agent.
Implement vector embeddings for semantic similarity search.
Model a simple decision problem as a Markov Decision Process.
Implement a basic symbolic planning system for an agent.
Train a simple RL agent to solve a decision-making problem.
Design a hybrid planning approach combining symbolic and learning-based methods.