ºÚÁϸ£ÀûÍø

AI-assisted coding, also known as agentic coding, is evolving rapidly. Since the introduction of Claude Sonnet 4.5 and comparable models, a new generation of AI tools has emerged that is mature enough for everyday use in software development.

While these tools offer tremendous opportunities for rapid experimentation and prototyping, building software for production environments requires a different approach. Current practice shows that relying on a single coding agent often results in errors, bugs, and inconsistent outcomes.

At the same time, recent developments have made it possible to build software reliably using multiple collaborating agents. By applying structured workflows with guardrails, quality controls, and validation mechanisms, organizations can benefit from agentic coding while maintaining high standards of quality and reliability.

In this masterclass, we will explore the current best practices in agentic coding. You will gain a practical framework for setting up an effective agentic coding pipeline and discover the most widely used open-source tools in this rapidly evolving field.

During this masterclass, you will:

• Understand the principles behind AI-assisted coding and agentic coding.
• Learn the difference between experimenting with coding agents and developing production-ready software.
• Understand why single-agent workflows often lead to errors and how these risks can be mitigated.
• Explore current best practices for software development using multiple collaborating agents.
• Learn how to apply structured workflows with guardrails, checks, and balances.
• Gain practical knowledge on setting up a robust agentic coding pipeline within your own development environment.
• Discover the most popular open-source tools currently available for agentic software development.

This masterclass is designed for Data Science and AI professionals who want to learn how to effectively leverage the latest generation of AI coding agents in their daily work. Whether you are working on data analytics, machine learning, AI solutions, software development, or innovation projects, this masterclass provides practical insights into both the opportunities and limitations of agentic coding.

The masterclass is particularly relevant for professionals who want to increase their productivity, improve the quality of AI-generated code, and learn how to establish reliable development workflows using multiple collaborating AI agents.

To get the most out of the hands-on sessions, participants should have some experience working with a development environment (IDE), such as Visual Studio Code, and feel comfortable using technical tools and the command line. Extensive software engineering experience is not required, but familiarity with programming and technical workflows will help you maximize the value of the masterclass.