6 POWERFUL AI Tools You MUST Start Using Immediately!

now regardless what kind of apps You're Building you should definitely be thinking about how to leverage AI in order to make your apps more robust now all of that used to be complicated and you needed to be technical in order to make it work but that is all part of the past because right now it's relatively easy and straightforward to a leverage AI to make your apps a lot better for your users and so in today's video I'm going to be showing you some amazing amazing tools that you can easily plug into your apps and leverage large language models to make your apps a lot more robust and also a lot more user friendly now before we get started as always if you're interested in taking your no code knowledge up a notch or two then you definitely need to join our amazing patreon community and you can learn more about our amazing patreon Community using the link in the the description below the video now it's also Worth to mention that the apps and the tools that we're going to be covering today are going to be either free or they're going to have a very very generous free plan that you can get started with without grabbing your wallet so the first tool that I want to cover today is called stack and it is available at stack Dash ai.com and this tool allows you to build and deploy AI applications in minutes the fastest and more reliable way to incorporate custom large language models llms such as chat GPT into your product or team alright so we're gonna go ahead and log in into my free plan right now so that I can show you what this tool does and how it works okay all right so here I am logged into stack and the first thing that we have here is the four tabs we have projects data fine tune and deployments so we can start a new project from or we can start from one of these templates right here so let me show you a couple of exam example so let's say you're looking to do something like a document database q a so you can simply click here and you can take a look at different templates that allow you to feed in data and do a q a on that data so you have document q a document analysis Google search URL air table notion API right so let's go ahead and try URL q a we're going to select this and as soon as we select the template we are redirected to this screen here we are on the right hand side we're seeing this kind of diagram here made up of components that you can drag and drop all over the place and then on the left hand side you have various things that you can kind of drag and drop into this app here and so just to show you briefly What's Happening Here is that we have an input here and this is where you can interact with the data okay and when we talk about data this is this this part right here right so you're specifying a URL that goes into embedding things and then we have the open AI which is our large language model here and here we can select maybe we want to use gpt4 and here we have a system prompt and we have the user prompt here okay and here is going to be the output that the model needs to mimic for us so what we can do here is we can click on play and the model just executed and this is the result we got the main character is Jerry Seinfeld okay so why exactly did it do that well it's all part of this prompt here right so it says here your helpful web assistant you will receive pieces of a document in the context and answer questions about it and the question in this case is who is the main character okay so we can change this and that's going to change the output so let's say we change the input to who are the characters okay now we're gonna click play and let's see the output that we're going to be getting and here the model finished executing this is what we have the context mentions several characters in the show Seinfeld including the principal characters for close friends and then at least the characters if you have seen Seinfeld let me know if this is the correct answer now if you're happy with this model you can click on publish right here and that's gonna publish your app here onto the internet means that you can access it using third-party applications using apis so if I go back to projects right here and click on deployments I'm seeing this project here called Web Scrapper QA and if I select curl here I have a sample code that essentially gives me access to this app this app that we just built using an API interface okay so basically if I'm building like a flutter Flow app and I need to do some work in the back end well that flutter Flow app can now access this particular app using this scroll interface now the flow that we've created here is barely scraping the top of the surface there's so many things that you can do there's Advanced use cases classifiers you can build chatbots document database classifiers all kinds of things where you can essentially feed various data to your llm and get the exact results that you are looking for so this makes it all very very easy to do now the next tool that I want to talk about is called relevance Ai and this is the fastest way to build supercharged AI tools take GPT to the next level and build end-to-end custom llm apps and agents that you can share use and run in book today fully managed low code and developer friendly okay so you can sign up for your free account which I've already done and so I'm going to go ahead and log into my free account right now all right so here I am logged into the dashboard right here and you can obviously create a new change from scratch or you can start with a chain via a template that already exists okay so you can do something like GPT on your files you can categorize text you can do GPT on my website you can ask internet to ask questions from Google Plus GPT so you can take a look at all of these templates and see if one of these templates is similar to the problem that you're trying to solve and if that's the case you don't really need to build the specific chain from scratch and so let's say you're looking to use an llm such as GPT on your own files okay so we're gonna click here use this template and we are going to start from a pre-made flow that we can customize according to our needs okay so this is called GPT on my files get your questions answered on any PDF CSV or audio files with GPT so the first thing that we need to do is we need to drag and drop a file so here I'm going to drag and drop a file and the file that I'm uploading is about Roman Empire this is a PDF that I create created from Wikipedia's page on the Roman Empire okay here you can enter a question so maybe you want to find something out about this specific file so I can say why did the Roman Empire collapse okay so we're gonna say run once right and here we are running our pre-made flow okay and this is the answer that we got the Roman Empire collapsed due to a combination of an internal and external factors and so the answer is very very concise but it is the right answer and so if you want to learn more about it right what you can do is we can clone this template and we can edit it we can see how it was built and so we've just cloned this sample and now we can learn more about it and tweak it if necessary and so the first thing that we have is we have this user input component here okay so this is the part where you are uploading a file that you want to learn more about it this could be a PDF CSV or audio files and you're going to have access to this variable here and then the next thing you need to do is you need to have an input field for the user to enter a question and this is what you're going to be given this is going to be the value right here and what's cool about this is that for the input field you don't need to have it as a text input field right you can have an options drop down on numerical input a text list Json you name it it could be a bunch of different things next we have the llm so if we expand that this is our LL lamp right so we're setting up the prompt here right this is the context which is the file text and this is the question that you know the user gets to ask in regards to the the document or the media or whatever they are interested in right and here we can specify the model so this is actually using GPT 3.5 16k we can choose a different model if you want and last but not least we have the chain output and this is where we're going to get the answer to our query okay so we can click on configure the output and as you can see we have steps prompt completion output answer this is the answer and this is going to be available as output.ans certain case we want to do something with that variable now if we're happy with the way that we built this flow we can go into use and we can run this in bulk okay so we can upload multiple files and ask different questions we can also create a chat bot we also have a schedule and then we also have API right so if we click on deploy here we're going to be given an API endpoint and we're also going to be showing a sample request body along with a sample curl request and you can easily access this flow using a third-party app so if you're building like a flutter Flow app and you already designed this flow here then you can easily connect your flutter Flow app using its API functionality to this flow right here which is awesome because now you don't have to worry about doing all of this llm stuff in your flutter Flow app you can easily build your app just to do the UI Style off and here you can worry about the llm stuff which makes the whole process a lot easier and more straightforward now the next tool that I want to show you is a tool that I'm very very excited about because you can build some interesting apps using this tool so this tool is called streamlit and it's available at streamlit.io and this is a tool that allows you a faster way to build and share data apps okay extremely turns data scripts into shareable web apps in minutes all in pure python no front end experience required okay now unlike the previous tools that I showed you this is going to be more of a low code type tool which means that you still need to write code but you don't need to you know get a PhD in computer science in order to build simple web apps and also there's a really nice benefit to using this tool that I'm going to share with you a little bit later in the video okay so so how exactly does this work well if you scroll down you can see that you write a little bit of code and you get beautiful web apps and so essentially this tool provides a very very rich framework for easily building all kinds of web apps okay so here if you go into Gallery you can see all kinds of different apps that you can build using this tool so there's a cheat sheet you have Goodreads analysis app and Goodreads if you're not familiar is like a book review type site where people review different books you have some road map you have a gallery a wave quick view you have a you know pretty map some kind of a mapping app extras here you have knowledge GPT you have e-charts image background remover and all kinds of very very interesting apps that you can build relatively quickly now since this video is focusing on AI let's go ahead and open up this app so that we can see what kind of apps that we can build quickly with this tool so this is what the app looks like and the first thing that we need to do is we need to paste our API key here so I'm going to go to my open AI API dashboard create a new key bullet stream copy this key right here paste it in here press enter and now it accepted our key so now what we can do is we can drag and drop a file here and then we can chat about this file so let me go ahead and upload the same file about Roman Empire okay so it's uploading the file it's indexing the document and it's done so now we can ask a question about the document so but why did the Roman Empire collapse and let's see if it works all right here's the answer the collapse of the Roman Empire can be attributed to various factors including the assimilation of Germanic peoples internal political conflicts and the disintegration of the Western Empire okay so it goes on and if you notice this answer is a lot more comprehensive than the answer that was generated in using one of the previous apps in the previous example and so it also provides the sources here as well and so as you can see this particular app works perfectly and so you can go back to the gallery here click on view source and you can take a look at how this particular app was built so if you choose knowledge GPT here which is the name of the app you can take a look at how this particular app was built and so as you can see this is the main file right here for the app and it's only 117 lines and so it's not a very very involved app to build it's not a very very complex app and that is because this streamlit framework provides a lot of the functionalities a lot of the tools that allow you to build all kinds of web apps very very quickly and easily and there's also a shortcut that allows you to build many of these apps in record time and that shortcut is getting gpt4 to help you build all kinds of apps so here's my GPT 4 conversation here and I'm going to make this conversation available so all I'm saying is write me a little app using streamlit we'll ask a user for input and then send it to gpt3 so very very simple app sure below is a simple streamlit app that has the user etc etc and this is the app that it created and so all you need to do is follow the instructions when it comes to building a brand new app on streamlit then take this piece of code and just copy and paste it in the exact spot where it tells you that you can put your code and as a result you're gonna have a very very simple app so just to show you how gpt4 is actually generating the right code for you if you take a look here you can see that we have this user input equals to sc.text area and this is that text area field that the user sees where they need to enter some input there right and so if you go to this knowledge app you have the same thing you have this St text area right and you have this you know kind of uh you know a hint as to what the user needs to do this this text that appears near the input so the user knows exactly what to do and it's the same type of code that gpt4 generates for us and the other thing about gpt4 is that it's really really good with python code you know it might not be as accurate with some of the other languages but when it comes to python it's has been very very solid and streamlit uses python as its language for building these web apps so you can easily ask gpt4 to write you a sample streamlit web app that you can copy and paste and a launch right away now the next tool that I want to show you is called cohere and this tool is available at cohere.com and here you can click on products and kind of learn more about this tool and so you can do summarizations you can do generate you can do embed semantic search where you rank classify and it has you know command models and embeddings as well okay so let's go ahead and log in that I can show you the power of this tool all right so here I am logged into coherence dashboard here okay and if you scroll down you have this quick start tutorials okay so you can sort customer questions you can summarize passages intent recognition analyze user sentiment and then you also have different ways of doing that including our favorite no code way which is this curl interface right here so now I can also click on models and here I can create a custom model that I'm going to create with my own data now the best way to understand the power of this tool is to scroll down and go to their playground so if I click on playground on I have different things that I can do okay so here's my generation Tab and examples of things that I can generate essentially is a LinkedIn post generator a Blog title Alternatives generator blog post title I can even see more and I have more options here if I go into classify I can classify different things right so here's an example of things that I can classify if I go into embed and what's going to happen is it's going to take a list of sentences and output a list of vectors allowing models to compare their meanings so let's say I go into generate here and let's say I want this blog post title and it says you generate five title with topic tone and audience and this is the input here okay write five titles for blog ideas for the keywords large language model or text generation okay so I'm gonna click on run right here and we're gonna wait for this generation to finish and so here's the result that we got we can obviously tweak it more change the input change some of the parameters now that's nice at all but one of the more powerful things that I can do is if you go into embed over here you can do some really really interesting things so if you click on see more you have a restaurant customer inquiries you have y combinator threat titles subreddit titles so if you click on y combinator threat titles uh we have our input here okay and that takes a list of sentences and outputs a list of vectors and here's our prompt identify five unique clusters from the data set of thread titles on on Hacker News and so if I click run here we get a really nice visual representation of different titles clustered together so here's one cluster right so it says here should I quit the field of software development did your life as a parent affect your life as a developer right so this is more kind of um live thing how do you deal with lack of motivation how are you getting through you know not burning out so this is kind of self-improvement right how do you create productive habits how do I they motivated to learn etc etc if you come over here we have three more clusters and they're kind of separate but yet they're still close together right so if you take a look at this who is Seeking a co-founder that your YC startup fail obtaining initial users for startup right and here we have interesting software books non-tech books which books mental models and then we have another one what is your favorite tech talk what's the best technical ebook so as you can see it's clustering it they're kind of similar but they're still individual clusters here we have another cluster the reading agent bring anything valuable what are some of the best hm comments that you've read etc etc and then last but not least this has to do with Google as you can see it's just me or that Google search recently get a lot worse someone hijacking Google Images what's up with Google etc etc so this is a really really nice kind of visual diagram that you're able to create with some data and here you can click on view code and you're going to be given a code that you can use in order to get this data back so if you're using python node go uh you can use these Snippets of code but typically if you're doing no call you're going to be using curl and so here's a full curl request that you can execute and get the exact results that we are seeing here so a really really powerful tool that allows you to analyze and kind of slice and dice uh your data in order to learn some interesting patterns that may exist hidden deep inside now the next tool that I want to talk about is called Monster API and this tool allows you to do something truly magical and that is to fine tune an llm to fine tune a large language model okay so generative AI Made Easy with monster API effortly access powerful generative AI models with our Auto scaling API AI zero management required okay so if you're not familiar what exactly is fine tune and how it helps you well you can think about it this way right you have an llm which is a generative AI model so it completes something that you may ask you to do but if you are looking to change the behavior you know maybe you're looking to teach it something new maybe teach it a brand new language or teach a specific terms or change the way that you wanted to respond to your queries then in that case you're not looking to give it embeddings you're not looking to give it new data you are looking to fine tune it because only fine tuning changes the actual behavior of an llm and this tool allows you to do just that so if you click on AI models so if you click on a models you can see all the models that it supports you can take a look at API Docs here you can learn more about fine tuning an llm so if you click here it walks you through the process you have a first second third and a final step right there and so if you're interested in more of an in-depth tutorial on this specific tool let me know in the comments below now last but definitely not least this tool is called metal okay and this is the llm developer platform for real world use from idea to production metal will help you build AI into your product so we're gonna go ahead and log in into my free account here and so here I am logged into Metals dashboard here and the first thing that we see is the three tabs here so we have data retrieval we have chat memory and we have settings so for data retrieval you're going to be seeing your indexes okay so you can create a new index but I already created an index here so if we click on this index I can add records that I want to have stored in this particular index so I can click on ADD records here and I can upload the file add text add image or use the API so what I'm going to do is add text over here and so here I can submit a specific piece of text and also attach metadata to it and so that way when I'm actually retrieving a specific piece of text I'll be able to get the metadata and that way I will know you know what the text is about and typically the way that you're going to be doing it is let's say you want to store a piece of text so for instance this paragraph right here you're going to copy this paragraph paste it in here and for the metadata you're going to be providing a Json formatted string that's going to have this text so I'm going to enter this here I'm going to close this I'm going to say index okay so now we've added another piece of text into the index if I reload this we now have three pieces of text stored in this index here now let's say I'm working with custom text well in that case I need to tell the system whether to pieces of text are related or not so in that case what I want to do is I want to click here enable this tuning mode right and then what I can do is I can click on this wrench icon right here and I can say label closer label further and that way I can fine tune specific pieces of text depending on my specific use case and here I can go into fine tuning here now once I finish labeling all of my text according to my specific preference I can click on this fine tuning area here and I will see all of my text here that I labeled myself okay so sell my label closer some are labeled further apart now when all of this is said and done I can click on run tuning and I can run a tuning job and that's going to create a fine-tuned version of all of these stacks of all of these embeddings that we created initially next you can click on clustering and you can generate clusters from your data that we submit it in our Index right here so lots of very interesting things that you can do and of course you'll be able to access all of this using an API as well now if you're watching this video and you're thinking to yourself you know I really really like the tools that you covered and I'm actually looking to leverage one or a couple or maybe all of these tools in my app that I'm building let's say with Florida flow or something like that which is the logical step well in that case what you need to do is you need to check out our amazing patreon community so if you're looking to level up your no code knowledge maybe you want to build a dream app that you've been thinking about for a long time or you want to become a freelancer or you want to get a job as a no-code developer then the content and the resources that you're going to discover inside of our private patreon Community is going to definitely help your no code development and accelerate your no code growth because when you become a member you're are going to get access to all the no code apps that are covered on this channel plus you're going to get access to extra content such as q and A's live streams behind the scenes content and our patreon supported masterclass Series where I do a deep dive and the topics that the community votes on and above and beyond when you become a member you're going to be supporting this Channel and supporting my work and that is highly highly appreciated so if any of that sounds like something that's going to provide you with value then you definitely owe it to yourself to check out our amazing community and consider becoming a member and you can take the first step by clicking on the link in the description below the video

Channel: James NoCode

Views: 4,042

Keywords: nocode ai tools, stack ai, relevance ai, metal.io, cohere ai, monster api, streamlit io

Id: vLoTPMec1Dk

Length: 26min 41sec (1601 seconds)

Published: Fri Aug 11 2023