FREE OpenAI Swarm + Crawl4AI AI Agent Beats CrewAI & AutoGen?🤖🔥 Multi Agent Orchestration Framework

What's going on everybody Josh pook here and in today's video we're going to be taking a deep dive look into swarm opening eyes new AI agent framework we're going to take a look and see how it actually compares to some of the leading Frameworks like crew Ai autogen and Lang graph and we're also going to see how we can integrate open AI new swarm with crawl 4 AI an open source framework that simplifies asynchronous web scraping and I'm even going to

Give you our strid swarm craw for AI marketing AI agent for 100% free in our GitHub repo which is going to include the craw for AI documentation as well as the Swarm documentation and examples so you can iterate on it and improve upon it and it already has five different custom agents built into it such as the user interface agent web scraper agent analyst agent campaign idea agent and copywriter agent where you can Implement any single website and it will scrape

All the data on the website generate a custom marketing plan complete campaign analysis marketing copy that you can use for advertising and much more and of course I'm going to provide you our in-depth document that will contain all the different links resources videos and any other additional information surrounding open AI swarm craw for AI and this custom agent so you can easily follow along with this video and just simply copy and paste to get this up and running as soon as

Possible this is going to be a good one let's dive right into it [Music] all right guys so I'm sure all of you have already heard about opening eyes new swarm framework I will briefly go over it as well as the blog they posted and then we'll dive into actually the AI agent with craw for AI and actually getting this implemented okay so here's the GitHub repo for swarm if you haven't already I would definitely recommend checking that out as well

As this blog post right here on their open AI cookbook you can see orchestrating agents routines and handoffs and this really just goes more in depth to the Swarm framework talking in-depth about routines as well as handoffs so I would definitely recommend checking out this cookbook those two links are going to help you if you want to go deeper into swarm okay so the key links are the Swarm repo of course the cookbook all right we also have the crawl for

AI repo right here if you're not familiar with crawl for AI it's a really good open- source tool for scraping and I've done videos on it before as well as the crawl for AI docks okay also of course is the stride swarm crawl for AI repo right here so that's this repo right here you're going to want to get clone it I'm going to show you that in just a second once we actually start setting this up and if you're not

Familiar with this repo guys I did a video yesterday announcing our new stride AI agents repo where I'm going to be uploading all the different AI agents we build on this channel here so this is going to be an extremely valuable resource so if you haven't already make sure you start it and watch it so you can stay up to date with these new AI agents okay and I would highly recommend checking out the first video right here if you haven't

Already it's a bit of a longer one but it is extremely valuable and I spent 7 Days creating it so definitely check that out as well as this in-depth 79 page document are surrounding AI rag agents as well as this document here that I talk about in the video and this is the master prompt engineering template all right so even if you're not using NN that video and these resources will be extremely helpful for you I can promise you that okay

So what is open AI swarm well open AI swarm is an experimental open source framework released by open aai for creat orchestrating and deploying multi-agent AI systems it's designed to make agent coordination lightweight highly controllable and easily testable so a few of the key features is its lightweight design of course as of now so focuses on Simplicity and ease of understanding it's open source it's built on opening eyes chat completions API not assistance so this allows for scalable and easily testable

Multi-agent systems now it's stateless architecture so it doesn't store State between calls similar the chat completions API which can limit its use in complex context dependent tasks now autogen crew Ai and Lang graph offers stateful operations allowing for memory retention between interactions all right so a few of the Core Concepts we briefly touched on them already but agents so this is the unit that performs specific tasks or functions then we have handoff so this allows agents to delegate tasks or transfer

Conversation control to other agents seamlessly and then routines so this is natural language sequences of steps defined with necessary tools for execution and here are a few different examples that opening ey provided in their GitHub repo right here are you tired of pouring thousands of dollars into appointment sets only to watch leads slip away imagine having a team of elite sales agents booking qualified appointments for you around the clock no more wasted time on training no more frustration with performance and

No more draining your budget on inconsistent and expensive call centers introducing stride agents AI powered appointment sets that work 24/7 never get tired and book appointments while you sleep trained on thousands of successful conversations our AI agents outperform human teams at just one tenth of the cost join the ranks of businesses that doubled their appointments and booking rates in just a matter of weeks don't get left behind in the AI Revolution visit stride agents.com now and transform your entire sales process

With Cutting Edge AI technology it's time to accelerate your stride with AI agents all right so the simple triage agent the weather agents the Airline customer service agent personal shopping agent and a bunch of other ones you can check out the GitHub repo there and lastly before we dive into this custom agent build we're going to talk about swarm versus crew AI versus autogen versus langra so swarm stands out for its Simplicity and lightweight design offering more granular control over agent

Interactions now swarm is like I said it's experimental so it's you know you know a lot of people have been saying oh it's only x amount of lines of code and blah blah blah blah blah which is definitely true these tools right here do have a lot more features and capabilities right now but swarm is still very new and if swarm just adds some of these features and capabilities then it definitely could be very competitive so autogen more sophisticated debugging tools

And logging functionalities crew AI focuses on rule-based agent design and structured workflows and then Lang graph support cycles and branching and workflows built in persist and then agent coordination so all four Frameworks excel in agent coordination which is a core feature of multi-agent systems however they approach it differently to open AI swarm uses the concept of handoffs to allow agents to transfer control based on context autogen and crew AI offers sophisticated agent interactions and rule-based design and then Lang graph provides

Support for complex workflows with cycles and branching so advantages with openingi swarm so flexibility modularity low learn curve at the moment and then the limitations so it's experimental no official support stateless and then way less features at the moment and then also a note on production grade usability autogen and cre AR are more mature and suitable for production use while open AI swarm is explicitly described as experimental and not intended for production experiments and then here is a table comparison between

Some of these tools here all right and here's an additional video I made on crawl for AI so you can check that out if you're new to the tool after this and then once again the first video that I mentioned before I would definitely recommend checking that out and these resources okay so let's get into this custom build and if you're not familiar with crawl for AI it's essentially a similar tool to something like fir crawl and I've also done a

Video on Fir crawl but it doesn't require an API CU so I really like this tool so if we go to the GitHub repo here in number two you'll see that we have a few different folders here we have the Swarm examples which is the Swarm examples and the re me so you can use that um say if you're using this within cursor you can use those examples as well as the crawl for AI docs right here which I uploaded for

You guys this is all the crawl for AI docs so if you want to iterate upon this improve upon this even if you don't know how to code you can input this into cursor use it as context and then improve upon this AI agent okay so you can go through read this look at some of the features and the Agents that we're using I'm going to Showcase them in just a second a few requirements python 3.10 plus that's a requirement with

Swarm and then the crawl for AI Library open AI API key you're going to need that and then the Swarm AI Library so first things first run this command to get clone the repo then second you're going to change directory into the folder and then you're going to want to create a virtual environment this is optional but recommended so here's an example run python DM virtual environment and then depending on if you're on Windows or Linux or Mac you're going to

Activate it using one of these commands right here to activate the virtual environment and then you're going to pip install requirements right here so if we take a look at what the requirements are that we're installing right now it is craw for AI all open AI python. EnV torch playright and then beautiful soup so once you install those requirements you're going to want to install the open AI swarm so you can run this Command right here tip install and then we're

Installing swarm or you if you have SSH access you can run this command and then you're going to want to set up your environment variables so you can either run this command to copy the EnV example to a newv file or you can just do that within your IDE but you're going to want to do that and then change out your open AI key right here and then save that as a EnV all right once you're done that you're going to

Want to run the python script all right now before I run the python script I'm going to quickly go over it a little bit not fully in depth but just briefly just so you can kind of see like I said you can check this out of course and change it around iterate upon it so we have some logging right here loading environment variables you can see the Imports up here we initializing our open AI client and then this is a function

To create the folder for the website so we're going to scrape um a website we're going to input a website URL it's going to scrape the content of it so this is the function to scrape the website it using crawl AI so scrape website using craw AI asynchronous crawler and then clean the content and then we're going to clean the content and structure the content with beautiful soup remove script and Style Elements right here get the text break it in so

This is all just um formatting right here then we're saving the clean content to a file and then we're saving the extracted content to a file and then this is a function to generate completions using open AI right here all right we have some different functions here here and then we have our different agents right here so we have the web scraper agent the analyst agent the campaign idea agent the copywriter agent the user interface agent all right and here you

Can see the different agent workflows their purposes their functionality of each individual agent and this whole agent isn't perfect by any means like I said you can definitely iterate on it improve upon it make it better and I would definitely change around some of these promps to get better output but this is just a starting point that you guys can use and get some good ideas from all right so now that we have all the requirements installed let's go ahead and

Test this so we're going to run python main.py all right so it's asking us for a URL to analyze I'm going to input stride agents.com as you can see scraping website warming up asynchronous web crawler is ready to crawl now it is started to crawl extracting content right you can also see here it has generated a new folder called stride agents we can see that we have the full website content right here this is all the different website content we can

See website content here this is just indexes of it and whenever you scrape a new website you'll see a folder is created with all the different markdown files in there all right so you can see we got our marketing analysis right here now it's asking for the target audience let's just say business owners to keep it simple marketing goals book more calls as you can see creating campaign idea generating completion for marketing strategist now we got a campaign idea MD file

Right here now it's generating completion for copywriter all right so now we got our marketing copy right here so in total we can see here that the demo has been completed we got our marketing plan MD file our website content our marketing copy our campaign idea our analysis and then our full website content so if we take a look here we can see analysis you know we can see all the different stuff right here we can see campaign idea right so

We can see different campaign ideas of course and the prompts that we're using in this AI agent definitely are not optimized um to my liking to what I would actually use this impr production this is just example I would definitely uh customize and improve upon these prompts to get better outputs okay same thing for marketing copy I would definitely make sure that you tailor it for your specific brand or for your specific use case all right so you don't just get

Cookie cutter AI generated stuff here um and same thing for the marketing plan but this should give you some hopefully some good ideas and be a really good starting point to for you guys to see what's actually possible with this as you can see here we got different campaign components personalized video messages inter interactive webinar series whatever the case may be and you can obviously get creative with this have different outputs that it generates have different agents and we're really not

Even U izing um all the different you know capabilities that the open ai ai swarm uh you know framework or whatever you want to call it has right now and I'm very excited to see um the continued updates that they put out for this you know even just where it's at right now even though you know I know a lot of people are saying that it's not comparable to something like crew Ai and autogen which I completely agree with right now

But us knowing open Ai and how quickly they could really turn things full blast and start pumping out new updates and shipping things if they wanted to uh I'm definitely confident in the future if they actually start taking this out of experimental mode that we will see some pretty cool things with this framework but other than that guys I just wanted to give you a quick video showing you a potential use case how you could integrate something like open AI swarm

With something very powerful like crawl for AI you could do this for many different Frameworks you could do this for fire CW you could do this for really anything the possibilities are endless but you guys can get creative with it let me me know in the comments what cool things you build with this and like I said all the links will be linked down below specifically to this GitHub repo as well as the document which outlines everything if you have any

Questions let me know down below and I'll try my best to help you out if you're new to the channel guys we upload videos every single day on AI automation Marketing sales Etc so if you like this type of content make sure to smash that like button and subscribe to stay up to date with the daily uploads also too guys if you like this video and you like AI agents I would definitely recommend checking out the video I made yesterday today

Like I said before as well as the different resources and master prompt template that I provide to you guys there for free I'm positive that you will get some good value out of it and if you guys haven't already joined our free community stric community.com I'll leave a link Down Below in the description our free Facebook group and Discord Channel you can join there and network with other like-minded AI entrepreneurs developers whatever the case may be also too guys if you

Run a business and you have leads or you need leads and you need them being set into your calendar or your sales team's calendar check out stride agents.com AI appointment CS with AI voice AI text AI email whatever the case may be that will book appointments directly into your calendar on autopilot while you sleep trained on thousands of successful conversational Pathways these AI agents will outperform real life humans at only a tenth of the cost guaranteed that sounds like something you're

Interested in check out stri agents.com or book a call with me down below and we could see if it would be a fit for your business let me know what you guys are going to build with with open ai's new swarm framework as well as if you're going to use crawl for AI or another use case another potential framework let me know in the comments down below what's the next AI agent that you want to see me build and give it

To you guys for free on our stride AI agents repo other than that guys I will see you in tomorrow's video keep hustling keep grinding and of course guys accelerate your stride take care