Скачать или смотреть How to build an AI Agent with Python (Agents, Tools, Guardrails)

In this video, we’ll build your first AI agent using Python and the OpenAI Agents SDK. You’ll learn how to set up a working AI agent from scratch, add guardrails to validate inputs, and even create a custom tool to save outputs. While the demo focuses on an interior design agent that generates room designs from a floor plan, everything you’ll learn can be applied to building any kind of agent.

We’ll walk step by step through environment setup, agent creation, input/output handling, and building a simple Streamlit interface to test the project.

GitHub Repository: https://github.com/IAmTomShaw/interio...

Chapters
00:00 – Intro: What we’re building
00:22 – Why the OpenAI Agents SDK
01:04 – Demo: Interior design agent
02:04 – Project diagram & workflow
03:02 – Setting up the Python project
04:49 – Creating a basic agent
06:00 – API key setup & environment variables
08:05 – First working agent (math example)
08:16 – Swapping to interior design agent
10:24 – Adding inputs (floor plan + design style)
13:27 – Using the image generation tool
15:25 – Displaying generated images
17:29 – Why guardrails matter
19:09 – Building the guardrail agent
24:13 – Testing guardrails with invalid input
28:42 – Creating a custom tool (save to database)
33:15 – Final results & improvements
34:39 – Adding a Streamlit frontend
35:19 – Wrapping up & GitHub link

If you enjoy the video, don’t forget to like, subscribe, and hit the bell so you don’t miss future videos! 🔔