Build a Django Coding Agent with OpenAI Tools | Prompt Engineering and Function Calling Tutorial

Learn how to build a Python-based coding agent that can plan, write, and edit code inside a Django application, powered by OpenAI function calling. In this hands-on session, you’ll see how to combine prompt engineering, tool definitions, and automated file system control to create a real working agent that builds and styles a full-stack app. What You'll Learn - How to build an AI coding agent that interacts with your local codebase - Difference between chatbots and agents (and why tools matter) - How to define custom tools in Python for reading, writing, and executing code - How to design effective system prompts that guide model reasoning - How to connect OpenAI API, Django, and Tailwind CSS - How to debug and iterate prompts to improve agent performance Technologies Used - Python, Django, Tailwind CSS, Font Awesome - OpenAI API with function calling - UV for environment and dependency management Join the next cohort of the AI Dev Tools Zoomcamp, our free online course that helps you use AI tools to write better code, faster: https://airtable.com/appJRFiWKHBgmEt70/shrpw7rk55Ewr1jCG Resources: - Workshop materials: https://github.com/alexeygrigorev/workshops/tree/main/coding-agent - Advanced workshop: https://maven.com/p/3cf911/build-your-own-coding-agent - AI Dev Tools course: https://github.com/DataTalksClub/ai-dev-tools-zoomcamp/ - Vibe coding workshop: https://www.youtube.com/watch?v=NSMXQk4Axig - Agents workshop: https://github.com/alexeygrigorev/rag-agents-workshop TIMESTAMPS 0:00 Workshop introduction and goal of building a Django coding agent 0:29 Course context and connection to the AI Dev Tools course 2:52 Demonstration of the Lovable tool and project inspiration 6:07 Explanation of the difference between agents and chatbots 11:14 Creation of a simple chatbot using the OpenAI client 13:33 Demonstration of tool use with the make_joke function 19:01 Setup of the development environment and installation of dependencies 24:45 Finalization of the Django template with Tailwind CSS and Font Awesome 35:06 Definition of coding agent tools for file system operations 40:04 Testing of file tools and implementation of file tree filtering 45:42 Configuration of the agent interface and tool description generation 46:36 Design of the system prompt for the coding agent 52:51 First execution attempt to build the To-Do list application 1:03:43 Successful creation of the styled To-Do list application 1:17:28 Exploration of alternative project ideas (Tic Tac Toe, Tomato Timer) 1:23:47 Summary of results and next steps for improving the agent 👋🏼 Support/inquiries If you want to support our community, use this link - https://github.com/sponsors/alexeygrigorev If you’re a company, reach us at [email protected] #AIAgent #Django #OpenAI #PromptEngineering #Python #FunctionCalling #AIDevTools #MachineLearning #AIWorkshop #CodingAgent #DjangoApp #ArtificialIntelligence #AITools #Developers #AIZoomcamp