Agentic AI Engineering Bootcamp: Practical Workflows, Best Practices, AI Coding, and Live System Building

This workshop is a practical, beginner-to-advanced introduction to modern AI engineering.
Most people have tried prompting LLMs. Far fewer know how to build useful, reliable AI systems in practice, or when not to use agents at all.
In the first half, we’ll cover the most useful concepts and decisions AI engineers need in real projects: when to use prompting, structured workflows, retrieval, or agents; how to choose the right architecture for the task; and how to avoid unnecessary complexity and hype. We’ll also share practical AI engineering best practices, including how we use tools like Claude Code and ChatGPT in real development workflows at Towards AI.
In the second half, we’ll build an end-to-end application live using AI coding tools. Together, we’ll create a system that takes a YouTube link, extracts and processes the transcript, translates it, and returns a clean frontend output with key bullet-point takeaways in just a few hours while sharing exactly how we use AI tools to augment ourselves as developers. The goal is not just to show a demo, but to teach a better way to build: fast, practical, agent-assisted, and grounded in real engineering judgment.
This workshop is designed to help participants develop intuition for choosing the right AI workflow, working effectively with coding agents, and building useful systems end to end.

Part 1: Practical AI Engineering Foundations
What AI engineers actually need to know today: core best practices, common mistakes, and how to think clearly about reliability, usefulness, and speed.
Part 2: Workflows vs Agents
How to decide between prompts, structured workflows, and agents; when agentic systems help; and when they add unnecessary complexity.
Part 3: AI-Assisted Development in Practice
How we use Claude Code, ChatGPT, and related tools in real engineering workflows to build faster and better.
Part 4: Live Agentic Coding Session
Building an end-to-end application live: YouTube link input, transcript extraction, translation, summarization, and frontend display.
Part 5: Wrap-Up and Practical Takeaways
Key lessons, recommended habits, and a framework for applying agentic AI engineering techniques in your own projects.

Requirements
Basic coding experience is helpful but not required either. No prior machine learning background is required.
Ideally, they’d install Claude code or an alternative (gemini CLI, cursor agents, Codex…) and have a Python environment set up just to ensure they have already done that once — but we can go over it with them too.

Target Audience
Developers, data scientists, ML engineers, product managers, CTOs, AI leads, and technical professionals who want to learn how to build practical AI systems and make better decisions about workflows, agents, and AI-assisted coding.