Course
AI Agents — From Theory to Living Systems
This course explores the evolution of AI agents, from fundamental theoretical models to their implementation in real-world, living systems. Participants will learn the principles of autonomous decision-making, interaction between agents, and integration with complex environments. The program combines theory, practical exercises, and case studies to prepare students for building intelligent multi-agent systems of the future.
Skills you'll gain
- test skill
Format
How the learning works
A structured path of video lessons, hands-on practice and mentor feedback.
Video lessons with theory and practice
Modules you go through at your own pace, with lifetime access.
7 modules
A clear step-by-step structure from basics to production.
Hands-on practice
Interactive tasks and homework on real cases.
Mentor homework review
Line-by-line feedback and unlimited resubmissions.
Certificate
A named certificate on completion.
Course program
7 modules · 0 lessons · 0 homeworks
What is an AI Agent and Why Do We Need It
You will learn what distinguishes an AI agent from ordinary software, explore the PEAS model, and study the main types of agents — from reactive to hybrid. In practice, you will create your first mini-agent in Python.
Agent Architecture and Mind
In this module, you will explore the internal architecture of AI agents — from the BDI model (Belief–Desire–Intention) to planning methods from STRIPS to LLM-based reasoning. We will dive into memory models, from short-term and long-term to modern vector databases like FAISS and Pinecone. You will also study Chain-of-Thought and self-reflection as the early forms of 'thinking'. In practice, you will build an agent with memory and a planner using LangChain and vector databases.
Language Models as the Core
In this module, you will explore large language models (LLMs) as the 'brains' of AI agents — their strengths and limitations. You will learn prompt engineering techniques to communicate with models effectively, as well as methods of self-critique and reflection that allow agents to improve themselves. We will also review advanced agent architectures like AutoGPT, BabyAGI, and MetaGPT. In practice, you will build your own 'mini-AutoGPT'.
Multi-Agent Worlds
In this module, you will dive into Multi-Agent Systems (MAS), where multiple agents interact, collaborate, and compete. You will study distributed planning, collective decision-making, and consensus-building. Case studies include Voyager (an agent exploring Minecraft) and CAMEL (dialogue agents). We will also discuss digital economies and the idea of 'agent colonies'. In practice, you will create two agents that collaborate and compete at the same time.
Learning and Evolution
This module explores how agents learn and evolve over time. You will study reinforcement learning with environments like OpenAI Gym, self-play and meta-learning, as well as evolutionary approaches where digital agents undergo natural selection. We will also discuss how to design agents that actually become smarter with experience. In practice, you will build an RL-agent that learns to play a game or optimize a task.
Ethics, Safety, and the Future
This module examines the ethical, safety, and social challenges of AI agents. You will discuss where to draw the line between a tool and a 'conversation partner', the role of filters and constraints, and questions of legal status for agents as 'digital subjects'. Finally, we will explore the future of agents as new social actors in diplomacy, science, and art. In practice, you will design a safe agent with built-in constraints.
Final Project: Build Your Own Agent
In the final project, you will design and implement your own AI agent that can remember, plan, and act. It should be integrated into a real environment (API, chat, or web service) and function as a useful tool or researcher rather than a toy. Example projects include: an agent-manager that organizes work in Trello and Slack, an agent-researcher that reads articles and generates insights, or an agent-therapist that works with dialogue and reflection.
Course authors
Learn from practicing specialists
Course video material
2
Овладейте кварцевыми резонаторамиИсследуйте преобразование энергии в кварцевых резонаторах. Углубитесь в механизмы создания стабильных частот и их применение.
Мастерство в Пульсоксиметрии: Полное РуководствоОвладейте методами пульсоксиметрии. Узнайте о сенсорах, LED технологиях и измерении кислорода для медицинских приложений.FAQ
sad
Test que
Still have questions?
Contact our team
We'll answer questions about the program, payment options and team training.
- Reply within one business day
- Team training from 3 people