Agentic AI Programming for Python Course

Source: https://training.talkpython.fm/courses/agentic-ai-programming-for-python

Tried using AI tools for programming, but tired of non-functional code, copying from chats, and endless “cleaning up” after artificial intelligence? There’s a better way.

“Agentic AI Programming for Python” will teach you how to work with agentic artificial intelligence—not just a chatbot or autocompletion, but a smart assistant that understands your code, runs tests, formats the project, and independently creates full-fledged functions under your guidance.

Under the guidance of Michael Kennedy, you will learn to use tools like Cursor and Claude to create real production applications—from scratch and based on existing systems. The course covers principles of code structuring, testing, error handling, and integrating AI into the development workflow.

You will understand how agentic AI differs from regular chatbots and how to collaborate with it as if working with a talented junior developer in your team, with clear boundaries, standards, and expectations.

What makes this course unique:

Focus on agentic AI, not chats – you will learn to work with AI that understands the context of your project and can act autonomously.
Real examples — analysis of functions and services actually working in Talk Python and Python Bytes products.
Customizing AI to your standards — how to train models to write clean, structured code with types, tests, and error handling.
Practice and control — master rules, commands, and scenarios for Cursor and Claude that make AI a true team member.
Support and development of legacy projects — learn to safely improve old code without accumulating technical debt.
Working with visual design — using screenshots and images to accurately convey the interface concept.
Safe experiments — Git approach to working with AI, allowing you to experiment freely and easily roll back.
Optimization of costs and models — how to choose suitable models and control usage and budget.

By the end of the course you will be able to:

Distinguish agentic AI from chat and autocomplete tools.
Customize Cursor and Claude for your projects and standards.
Plan and implement complex features step by step in collaboration with AI.
Build comprehensive Python applications: CLI, web services, and production features.
Work with context, documentation, and agents that self-correct errors.
Create tests, set up logging, and apply modern Python techniques (async/await, typing, etc.).
Use AI as a productivity accelerator, not a source of chaos.

Who this course is for:

Python developers disillusioned with chat AI tools.
Professional engineers seeking to increase productivity without losing code quality.
Team leads and architects looking to integrate AI tools into their team’s workflows.
Indie developers and freelancers aiming to implement ideas and MVPs faster.
Developers maintaining legacy projects and those who want to add new features without technical debt.
AI skeptics who want to see how a properly configured agent can genuinely aid in development.