Module [Module Number]	Agentic Artificial Intelligence [1277BSWI12]
Regular Cycle	Winter Term
Teaching Form	Lecture & Exercise
Examination Form	Software project, Written report, Presentation
Teaching Language	English
ECTS	6
Instructors	Tim Alvaro Ockenga
KLIPS
Syllabus

 

Content

This course delves into the next frontier of artificial intelligence: autonomous AI Agents. Moving beyond simple AI models that predict or classify, we will explore intelligent systems that can perceive their environment, reason, make decisions, and act to achieve complex goals. As AI continues to evolve from a tool for analysis into an active partner in digital processes, understanding AI Agents is becoming critical for future business and IT leaders.

We will trace the evolution of AI Agents from their theoretical foundations in computer science to the current state-of-the-art, with a special focus on the paradigm shift introduced by Large Language Models (LLMs). You will learn how LLMs act as the "brain" for modern agents, enabling sophisticated reasoning, planning, and tool-use capabilities that were previously unattainable. The course will bridge the gap between scientific literature and practical application, equipping you with the knowledge to strategically leverage these technologies.

Key topics include:
- Foundations and Evolution: From symbolic AI to LLM-driven architectures.
- Core Components of Modern Agents: Understanding perception, memory, planning, and action loops (e.g., ReAct).
- The Role of Large Language Models (LLMs): How LLMs serve as the core reasoning engine for agents.
- Agentic Design Patterns: Developing agents that can use tools, access APIs, and interact with external systems.
- Managerial Implications & Strategic Value: Identifying high-impact use cases, assessing ROI, and managing the integration of AI Agents into business processes.
- Ethical and Governance Challenges: Discussing the risks and responsibilities of deploying autonomous systems in real-world scenarios.

A central component of this course is the hands-on exercise. You will work in teams to design, develop, and deploy your own LLM-based AI Agent to solve a practical business problem, giving you direct experience with cutting-edge frameworks and technologies.

Requirements

While no prior expertise in machine learning is required, a high level of interest in modern AI technology is essential. Basic programming skills (preferably in Python) are highly advisable, as they will be crucial for the practical exercises. You should be motivated to engage with both conceptual literature and hands-on technical development.

In this context, it would be perfect if you to have taken and know the topics covered in the course "Analytics and Applications" (14277.0605).

 

Learning Objectives

Upon successful completion of this course, students will be able to:
- Understand and explain the fundamental concepts, architectures, and historical evolution of AI Agents.
- Analyze the capabilities and limitations of modern LLM-based agents and differentiate between various agentic designs.
- Design and develop a functional prototype of an AI Agent for a business-oriented task using current programming frameworks.
- Identify and evaluate potential use cases for AI Agents within an organization and articulate their strategic business value.
- Critically assess the managerial, ethical, and operational challenges associated with deploying autonomous AI systems.
- Effectively communicate the design, function, and implications of an AI Agent to both technical and managerial audiences.

Timeline

We hold (the double amount) three 90-minutes sessions weekly (lectures & exercises) but only half the semester. Our goal is to provide you with the necessary skill to work on your team projects as fast as possible. After this you will work on developing your own agentic AI system the rest of the semester. You will have to submit your final work and report at the end of the semester (Vorlesungszeit).

Assessment

The course grading is threefold:

• Repository
• Written report
• Final presentation