Build an AI Agent with AI-Assisted Coding

Description

Our Q&A will happen on Discord so come join us there and ask questions live (#podcasts channel)! https://discord.gg/jmZHZApU In this free live workshop with Isaac Flath (Kentro Tech, ex-Answer.AI), Eleanor Berger (OKIGU, ex-Microsoft, ex-Google), and Hugo Bowne-Anderson (Vanishing Gradients), you’ll learn how to use AI coding tools to design, prototype, and build an AI agent, end to end. Most developers use Copilot or ChatGPT to autocomplete lines of code. In this hands-on session, you’ll learn to go beyond autocomplete and use AI as a genuine collaborator, one that helps you reason, plan, and build systems that actually work. You’ll learn how to: - Break down an agent problem into components AI tools can code - Prompt and guide your AI Assistant Cursor to implement each part (we’ll provide instructions for whatever assistant you use: Cursor, Claude Code, and more!) - Integrate generated code into a working Python application Iterate, debug, and improve your agent collaboratively with AI We’ll build a working agent together, a simple retrieval-based assistant that can: - Call external APIs - Retrieve and summarize information - Handle user input interactively Along the way, you’ll see how to: 🧠 Use AI as a design partner, not just a code generator ⚙️ Evaluate and correct model-generated code 🔍 Debug collaboratively with AI tools 🔁 Apply iterative development principles to AI-assisted workflows By the end, you’ll understand how to use AI coding assistants to go from idea to functioning agent, while maintaining control over quality, logic, and architecture. 00:00 Introduction and Welcome 00:20 Meet the Hosts and Workshop Overview 04:34 Workshop Goals and GitHub Link 06:22 Agentic Coding and Project Setup 09:09 Detailed Project Instructions 13:15 Implementation Steps and Agent Tasks 22:54 Planning and Research with AI 26:16 Executing the Plan and Model Selection 34:56 Reviewing the Plan and Adjustments 45:01 Finalizing the Plan and Starting Implementation 50:00 Context Windows and Effective Usage 50:52 GitHub Tools and Context Management 51:49 Executing the Plan: Step-by-Step 55:06 Multi-Agent System Implementation 56:30 Using Pydantic AI for Summarization 01:15:01 Continuous AI and Remote Agents 01:21:23 Debugging and Observability with Logfire 01:32:14 Final Thoughts and Course Information