Most developers already use LLMs in their work, but they tend to stop at chatbots and don’t fully understand how autonomous agents actually work. This webinar will help you get to grips with AI agents and learn how to apply them in your work as a developer.
AI agents are everywhere these days—in articles, at conferences, and in job postings. But when it comes to putting theory into practice, questions arise: what makes a system an “agent,” how can you allow AI to safely call your APIs, and how can you avoid spending twice as many tokens as necessary? Most online resources are either very superficial or dive straight into details without any context.
In 90 minutes, you’ll get a structured introduction to agent architecture, watch live coding from the first line of code to a working example—and go home with a Starter Kit for your own experiments.
Agenda
Block 1. From Chatbots to Agents What makes an LLM an “agent.” The difference between a chain and an agent. When you need an agent, and when a simple prompt is enough.
Block 2. Anatomy of an Agent Architecture: Planning, Memory, Tools. The ReAct (Reason + Act) pattern. The “Planning — Action — Observation” cycle.
Block 3. Live Demo Hello Agent: what a simple agent looks like. Demonstration of tool calling. Integration with an external API. Real-time debugging.
Block 4. Ecosystem and Limitations Common issues: loops, excessive costs, debugging challenges. Basic security: sandboxing and execution control. Framework overview: LangGraph, CrewAI, PydanticAI, the MCP concept.
Block 5. Q&A and Roadmap Answers to technical questions. Next steps and resources for self-study.
Who will benefit
Junior and Middle developers — if you’re already using LLMs for coding and want to understand how agent systems work from the inside.
Full-stack developers — if you’re interested in automating business logic with AI: user support, analytics, and request processing.
Python and JavaScript developers — if you want to understand industry standards and figure out which direction to take in developing your AI skills.
Tech Leads and Architects — if you’re evaluating the feasibility of implementing AI agents into a product and want a basic technical understanding and an overview of the risks.
Speaker
Viktoria Dvornik, Senior Data Scientist at N-iX, with experience in Python, ML, and cloud infrastructure (GCP, Azure)
Participant Requirements
— Basic experience in Python development
— Understanding of REST APIs and data formats (JSON)
— You are from Ukraine and currently in Ukraine
— Willingness to provide public feedback after the webinar
— 90 minutes available on June 25
It would be helpful, but not required: Python and a code editor installed—if you want to start experimenting with the code right after the webinar.
Register and get the Starter Kit: basic agent code and architecture diagrams for your own experiments.