FREE RAGapp AI Agent RAG Engine Beats ChatGPT?🤖 Open Source PDF RAG Web Search Vector Embedding Code

What's going on guys Josh pook here and on this channel we've covered many different ways of implementing rag AI we've covered using it with no code Builder such as NN we've covered chat rag tools we've covered implementing it with graph rag uh flow wise Lang flow a bunch of different options and in today's video we are going to cover one of the simplest ways the easiest ways you can start using a gentic rag in any Enterprise this tool is called rag

App I'm going to show you how you can set it up it is 100% open source and very easy to get started with let's dive right into [Music] it all right guys so rag app is exactly what it says in the name it's a rag app and it is a simple configuration of open AIS custom gpts there's also other um a uh llm models that you can use and it's Deployable in your own cloud infrastructure using Docker built using llama index

All right so it's very easy to get set up like I mentioned you simply just run this Command right here I'll show you that in just a second and the cool thing is once we get it set up there is three main endpoints there's the admin UI which will go over the chat UI and the API endpoint which allows you to actually use this um use the API endpoints and you could even use them in your own apps applications as you're

Building things so pretty cool stuff we're not not going to go over the whole read me section right here but you can go ahead check it out if need be if you want to learn more about the rag app now like I mentioned it is super easy to get going with this thing you're just going to run this simple Docker Command right here now before you do this of course you're going to need Docker installed and running on your system so

If you don't have that already go to doer. and download it and make sure it's open on your system and then run that command once you run that that is going to set all the different endpoints up as you can see right here I got it running right now and for example if I go to the admin UI I can see the admin UI is right here we'll go over this in a second the chat UI is right here and then

The backend API uh UI is right here so we can see all the different API endpoints right here we have the chat endpoints the sandbox management endpoints chat config model config agent knowledge base llama Cloud a bunch of different endpoints right here the open API so um you can use that if you're building applications or whatnot all right so the main thing you're going to want to use to start off is the admin uh UI all right so this is going

To allow you to configure different things such as agents knowledge or test the chat below and then once you're satisfied you can start the app or use the API so once you're done building this out you can actually just go over here to the chat and use it or you can use a chat in here like I said but uh first things first we can add an agent so you can click on the agent dropdown and you'll be able to see

That you can add different agents here so you can create an agent right here select a specific template there's templates right here we have an assistant agent a financial analyst agent we have a researcher agent a reporter agent okay you can see there's one right here uh as an assistant we got a backstory right here so uh expert in answering questions using the provided knowledge you can see this is just the background story of the agent the goal of the agent

So the primary OB objective right here and then we can see if we want to use a custom system prompt so Ur rooll backstory your personal goal is goal whatever you want to put in here and this prompt is used to generate a system prompt for the agent and yeah pretty self-explanatory then we have tools so this is pretty cool as well you can of course use knowledge now this will obviously use rag so it's going to use the embedding models

Actually once you get this um once you first open the app it's going to ask you to provide your open AP open AI API key right here um so you can use open AI Gemma AMA too so you can use local model any local model you want which is really nice um you can use mistal Gro Azure t- systems Etc now we're going to use open AI just cuz it's the simplest and we're going to use GPT 40 and then we're

Just going to use the test uh text embedding uh small right here and yeah if you were to use a llama of course you just put it to a llama and then just click update okay so that is how the used knowledge rag system is working from U your embedding model of course you can also enable web search so this will allow to find information on the internet searching duck Dogo you can enable Wikipedia you could enable document generator so generate

A document file PDF HTML so let's enable that you can enable image generators so generate images from provided text using the stability AI API so this you would have to put an a stability AI API key we're not going to do that right now but if you do want to generate images of course you can just go ahead and put that in and then we have open API so make requests to external API using the information from open API specs so

This would be the URL to the spec right here I'm not going to do that right now and then we have code interpreter so execute python code in a sandbox environment using e2b Code interpreter so we've covered e2b on this channel um many time or a couple few times about three uh two or three times um but basic basically if you want to use this you would just go to e2b dodev and get an API key right here you can get

One for free I'm not going to bother using it right this second but uh you could go ahead and use this and then code uh artifact so if you click here actually you know what since both of these do require e2b I will go ahead and input my key right here all right so once you create an e2b account you would simply just come to keys right here you don't have an API key you could add one right here and then

You would simply just copy this API key like so then you would go go back over here and you would paste it into here so now I'm pasting this API key in here and we can select the code interpreter as well as the artifacts like so okay now that we got our agent SL assistant made we can actually go ahead and add our knowledge so manage your own knowledge data in chat um to chat with you can consider using llac Cloud

For hosting your data so if you want to use that you could you connect to llac Cloud um set up the index name project name and add your API key and you could also just upload files like so so okay so I'm going to add this a2p 10 DLC registration bulletproof guide for go high level that I made I did a video on this I'm going to add this PDF right here right we can see that it is uploading are you

Tired of pouring thousands of dollars into appointment Setters 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 Setters that work 24/7 never get tired and book appointments while you sleep trained on thousands of successful conversations our AI agents outperformed human teams

At just oneth 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 cuttingedge AI technology it's time to accelerate your stride with AI agent going to also add this vs code PDF doc which has like some top vs code extensions so I'll be able to ask it what are the best extensions so

We got this uh these two test documents right here you could use llama parse you would just need to put your llama Cloud API key and you could use rerank right here you would just need a coh here API key not going to go ahead and use those right now but it is nice that they have all these different options um which is actually really cool because um this tool has been around for you know not a long time but um

A decent you know over a few months and um they have been making improvements and upgrades like before it did not have all these customizations which is really nice that they have these now and then we can go to chat config right here so we can add like suggested conversation questions like what is a2p okay we could add another one such as what are the top VSS code extensions at and suggest next question whether are two suggest the next questions to

The user based on the conversation so you could do this um in lining text citations whether to cite the text uh cite the text in the response so you could do this as well to site the citations of course similar to like how perplexity would work or anything like that search GPT okay and now that we have that all set up uh I don't know if we need to click update no that's just for the model okay I think we are

Good to go let's go ahead and test something out here so actually let's go to the front end app here see if we refresh it now we can see that there we have these two prompts right here what is a2p and what are the top DS code extension so if I click here you can see that it is generating a response right now okay so ATP or application to person is a messaging system that enables businesses to send messages to individuals

Using 10 digigit long code 10 DLC right blah blah blah blah blah you can see this we can see the sources right here and it's giving us follow-up questions how does a2p messaging differ from P2P okay let's go ahead and ask that and I really like the nice clean UI interface it's actually really nice to be to be honest it's very nice and simp simple right um as you can see it's giving us inep some indepth answers right here it has

Like formatting with Bolding and everything I actually really like how it has these uh questions afterwards that you can ask it now I'm just going to go ahead and refresh this and let's ask it what are the top vs code extensions and also to you could see that it was pulling from that that other um okay so this looks like it's actually using it looks like cuz the question we asked is definitely a question that uh you know could be used

From our rag or from the web so it's actually pulling from the web here which is actually using the other tool that we had in our agent which was the web search which is actually really nice to it use that um because that shows us some other tests so we can actually click here show events and it's saying call tool right here and we can see it called duck Dogo use the query right here I'm going to say no use the

Knowledge so I'm saying no use the knowledge base uh for what the PDF recommends and then I'm going to send that right and now we can see that is using the source right here you can see show events and it is calling the query Engine with inputs right here retrieving context for top vs code extensions retrieve three sources using uh to use context for query then we got python extension pack GitHub copilot path intelligence uh GitHub poll request better comments rough

We got material icon theme Adam one dark theme got the sources right here you can click here and it shows on page what where like that source is coming from Source number two right here hover okay so you can see all the different ones on page five right here it's really nice that it shows that we can ask followup questions like how do I install these extensions can you recommend any uh extensions for JavaScript development like if I click this I

Don't think there's any in the PDF so I think it actually just okay I think this was just based on using no tools now all right now I'm saying search Elon Musk on Wikipedia and tell me info so let's go ahead and run that now they should use the Wikipedia tool let's see the show events here so this is calling let's see so calling tool load data with inputs page Elon Musk so right here here um I think this used the

The I don't know what specific tool it used here but it did use something uh and then we're giving some information on Elon Musk right here you know all the information on his background um it's giving us some followup questions as well so all in all I really like this rag tool in terms of its Simplicity it's super easy to get set up with it's one of the most easy tools you can use that I've found so far that we've covered

On this channel we've covered other open source rag tools we've covered verba we've covered a few different ones you know ones that allow you to chat with your PDFs or documents or whatnot um but this one's really nice because it lets you create these agents really easily you get access to these pre-built tools and functions you can even add like image generator and these other ones with other services which is really nice um code interpreter code artifacts so let's actually try

Out the code artifacts real quick generate me some code for python uh for the snake game and run it okay so this one is definitely taking a lot longer as we can see the tools are calling right here so it's calling artifacts inputs generate python game with pi game um you got the inputs right here and calling tool as well so just probably tooken at least a minute so far let's go ahead and wait all right it's been about four or

Five minutes I'll give it a few more minutes if it doesn't generate I'll try one more time and if it still doesn't generate then I don't know what to say okay so for some reason this is just taking like it's been about 5 to 10 minutes now and all right guys so I'm having issues with the E2 b side of things if you guys were able to get this to work or maybe I was doing something wrong let me know in

The comments down below to help out others but everything else seemed to work pretty well the web search the use uh knowledge base the Wikipedia um also too guys the document generation which I just actually tested out so I just said generate me a document of uh the basics of a2p with a PDF so we can go ahead click on this see the document right here it's actually really nice that it was able to do this okay so this is pretty

Cool um this actually really nice for uh formatted and um we could go ahead download this also too guys if I didn't realize if you click on these documents you'll actually be able to see them in the side view which is actually really nice as well too um this something I didn't even realize until now so all in all I really like the UI of this tool it's really simple really easy to use and it actually just works pretty much uh

Other than the e2p um I'm personally having issues with I'm sure some of you will probably be able to get it to work okay can write code um but yeah I was just having some issues with the e2p and they have the image generator the open API knowledge works really well and then you configure the chat uh and everything like that and then like I said too if you do want to implement this into one of your custom applications or whatnot

You can go ahead mess around with the API right here all in all guys that's pretty much it for rag app let me know what your thoughts are about this in the comments down below let me know how you think it compares to some of the other rag tools we covered on this channel some of the more complex ones as well as the different uses of rag um in your day-to-day in your you know Enterprise or business but if you guys

Are new to the channel we upload videos every single day on AI Business Marketing sales automation AI agents whatever the case may be so if you like that type of content you got some value here make sure to comment down below like the video and subscribe to stay up to date with the daily uploads thank you guys so much for the recent 11k subscribers road to 20K on the way also too guys if you haven't already join our free community stri

Community.com I'll leave a link down below for our free Facebook group and Discord channel so you can join that and network with like-minded individuals and lastly if you guys run a business and you need help implementing AI agency into your business whether that's Ai call centers AI appointment cers or building out your entire CRM processes systems marketing process lound prospecting running PID traffic media buying whatever the case may be look a call down below to speak to myself and my team

And we can see if there's a potential fit or not 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