I want to build AI Agents

@qbuilder

11.2K followers · usually 1.2K views a post

Views

4.1K

Likes

238

Comments

29

Shares

35

SoundTikTok sound · original sound - qbuilder

Promoted product

Agentic AI course

Why it worked

The video provides a clear, actionable list of resources for a trending topic (AI agents), presented by a knowledgeable creator. The creator's approach of offering direct links via DM incentivizes engagement and community building.

Summary

This video recommends three resources for individuals looking to build AI agents. It highlights an essay by Anthropic, Andrew Ng's Agentic AI course on DeepLearning.AI, and the Claude cookbooks on Github as essential tools for learning and development.

Structure

1Introduction to building AI agents
2Resource 1: Anthropic's essay
3Resource 2: Andrew Ng's Agentic AI course
4Resource 3: Claude cookbooks
5How to use these resources

Original caption

I get asked every day where to start with AI agents. These are the only 3 resources you actually need 👇🏻 One: Anthropic’s Building Effective Agents essay, it’s short, free, written by the people who build the models and here is how you learn to think about agents before writing any code. Two: Andrew Ng’s Agentic AI course on DeepLearning.AI. The four design patterns that power every agent (reflection, tool use, planning, multi-agent collaboration) taught in raw Python, it’s free and you can start today. Three: The Claude Cookbooks on GitHub. The reference implementation for Anthropic’s paper plus working code for RAG, tool use, and MCP, this is a toolbox you come back to every time you get stuck. Starting point should be reading the essay, take the course and use the cookbook. Comment AGENTS and I’ll DM you the links 🙌 #ai #Tech #developer #claudeai #code

More from @qbuilder

Product placement

Anthropic: ACTS - it is the company that builds the models and publishes the essay on building effective agents. Removing it would change the context of the video. Claude: ACTS - it is an AI model that the Claude cookbooks are designed to help developers build with. Removing it would change the context of the video. DeepLearning.AI: ACTS - it is the platform where Andrew Ng's Agentic AI course is hosted. Removing it would change the context of the video. Andrew Ng: ACTS - he is the instructor of the Agentic AI course. Removing him would change the context of the video. Github: ACTS - it is the platform where the Claude cookbooks are hosted. Removing it would change the context of the video.

Call to action

DM me three links

On-screen text

I want to build AI Agents BUT I dont know where to start 3 resources YOU need 1. Anthropic's building effective agents is the other course out there it explains actually think about agents When and how to use agents Building blocks, workflows, and agents Building block: The augmented LLM When (and when not) to use agents The prompt chaining workflow Workflow: Routing core patterns Why I am here today Product Building effective agents Erik Schliuntz Barry Zhang HP LoC here before any code 2. Agentic AI course is Andrew Ng's Agentic on Deeplearning AI teaches design out there about tool use and agent collaboration Instructor Andrew Ng Enroll Now 3. Claude cookbooks on Github where you to see representation plus for rag MCP full agents a course a toolbox images managed_agents misc multi-modal observability pattern_agents scripts skills tests third_party tool_eval tool_use api_example README.md Building Effective Agents Cookbook Reference implementation for Building Effective Agents by Erik Schliuntz and Barry Zhang This repository contains example minimal implementations of common agent workflows: Basic Building Blocks Prompt Chaining Routing Multi-LLM Parallelization solves start I would anthropic essay agents next to build them and Claude cookbooks get Agents comment agents three links

Transcript

I get asked every day where to start with AI agents. These are the only 3 resources you actually need 👇🏻 One: Anthropic’s Building Effective Agents essay, it’s short, free, written by the people who build the models and here is how you learn to think about agents before writing any code. Two: Andrew Ng’s Agentic AI course on DeepLearning.AI. The four design patterns that power every agent (reflection, tool use, planning, multi-agent collaboration) taught in raw Python, it’s free and you can start today. Three: Claude cookbooks on Github where you can see real working code. It has the reference implementation for Anthropic paper plus examples for rag and MCP, full agents. It's not a course, it's a toolbox. Every time you hit a wall building your own agent, you come here and find a pattern that solves it. So if I would start from the beginning, I would read the Anthropic essay to understand agents. And next Andrew Ng's course to build them. And at the same time use the Claude cookbooks when you get stuck. Comment agents below and I'll DM you all three links.