<b>Flowwise AI is a tool that makes it</b> <b>incredibly easy to create powerful AI</b> <b>applications without</b> <b>writing a single line of code. It is open</b> <b>source, free to use, and was</b> <b>developed using the extremely</b> <b>powerful LangChain framework. If you are</b> <b>unfamiliar with LangChain, it is a</b> <b>development framework that</b> <b>allows developers to build AI-powered</b> <b>applications and has quickly grown in</b> <b>popularity.
And Flowwise</b> <b>makes all the features of LangChain</b> <b>available via an easy-to-use</b> <b>drag-and-drop user interface. </b> <b>So instead of having to write complicated</b> <b>code, all you have to do is drag-and-drop</b> <b>these nodes onto a canvas in order to</b> <b>create some complex AI applications. </b> <b>Flowwise has become my favorite platform</b> <b>for creating AI chatbots</b> <b>and agents for my clients,</b> <b>and I can tell you from experience that</b> <b>it's possible to create</b> <b>some impressive AI applications</b> <b>using Flowwise.
It's also easy to</b> <b>integrate Flowwise into</b> <b>our websites or applications</b> <b>using either the embedding or API</b> <b>endpoints. Let's start by</b> <b>setting up Flowwise locally. </b> <b>We will have a look at deploying Flowwise</b> <b>to the cloud in this series,</b> <b>but it's always a good idea</b> <b>to start off with a local instance for</b> <b>development and testing.
Setup is</b> <b>extremely easy, and we can</b> <b>see on the Flowwise website that it</b> <b>basically involves calling</b> <b>two commands in the terminal. </b> <b>This will install Flowwise globally on</b> <b>our machines, which I'm</b> <b>personally not a fan of,</b> <b>so I will also show you two more methods</b> <b>for installing</b> <b>Flowwise in a specific folder. </b> <b>And if you stick around, I will also show</b> <b>you how you can use a</b> <b>simple script that I wrote</b> <b>specifically for this series to quickly</b> <b>install and update</b> <b>Flowwise.
In order to use Flowwise,</b> <b>you do need Node. js installed, so go over</b> <b>to nodejs. org and install</b> <b>the LTS version of Node.
</b> <b>Then open up your command prompt or</b> <b>terminal, and then enter the</b> <b>command npm install-g flowwise,</b> <b>and press enter. This will install</b> <b>Flowwise globally on your</b> <b>machine. If you would like to</b> <b>install Flowwise in a specific folder</b> <b>instead, then skip ahead to the next</b> <b>method.
For the majority</b> <b>of you, this method will be perfectly</b> <b>fine. After installation has completed,</b> <b>we can start Flowwise</b> <b>by executing npx flowwise start. This</b> <b>will start a local server, and we can</b> <b>access Flowwise by going</b> <b>to localhost 3000, and you should see a</b> <b>dashboard like this.
So</b> <b>whenever you want to start Flowwise,</b> <b>open your terminal, enter npx flowwise</b> <b>start, and you can then</b> <b>access Flowwise by going to</b> <b>localhost 3000. In order to update your</b> <b>Flowwise instance, you can</b> <b>simply run npm update-g flowwise,</b> <b>and this will update your Flowwise</b> <b>instance to the latest</b> <b>release. And if you ever wanted to</b> <b>uninstall Flowwise, you could simply run</b> <b>this command.
Now for</b> <b>the majority of you,</b> <b>this installation process will work just</b> <b>fine, but if you wanted to install</b> <b>Flowwise into a specific</b> <b>folder, we can do that too. Installing</b> <b>Flowwise into its own folder</b> <b>actually opens up a few more</b> <b>advanced features, which we will have a</b> <b>look at during the course of</b> <b>this series. This is not too</b> <b>complicated, but it does involve a few</b> <b>steps, and after this I will</b> <b>show you how you can download a</b> <b>single script that will run all of these</b> <b>steps for you in one single</b> <b>go.
For this method, you do need</b> <b>get installed for your operating system. </b> <b>So download get from</b> <b>this website and install it. </b> <b>Then go over to the Flowwise repository.
</b> <b>I will leave a link in the</b> <b>description of this video,</b> <b>or you can access it from the Flowwise AI</b> <b>website by clicking on</b> <b>GitHub. First, click on code,</b> <b>then copy this URL, then create a new</b> <b>folder on your machine, and</b> <b>open up the command prompt. In</b> <b>Windows, we can do that by clicking on</b> <b>this address bar and</b> <b>entering CMD.
Then enter get</b> <b>clone and paste in that URL, and this</b> <b>will create a Flowwise</b> <b>subfolder, and if we open this up,</b> <b>we can see all the code that makes up</b> <b>Flowwise. Then open the</b> <b>command prompt in this folder,</b> <b>so I'll simply enter CD, Flowwise, and</b> <b>Enter. Now we need to install the</b> <b>dependencies for the</b> <b>Flowwise project, and we can do that by</b> <b>running yarn install.
If</b> <b>you run this and get an error</b> <b>message, something in the lines of yarn</b> <b>is not available, you can</b> <b>easily install yarn by running</b> <b>npm install --global yarn, and you should</b> <b>then be able to run yarn</b> <b>install. This could take a few</b> <b>minutes to complete. After installing the</b> <b>dependencies, we just</b> <b>have to run one more</b> <b>command, and that is yarn build, and we</b> <b>can now finally start</b> <b>Flowwise by running yarn start,</b> <b>and this will spin up the local server,</b> <b>which we can access from</b> <b>localhost 3000, and if I open up</b> <b>localhost 3000, I'm able to access</b> <b>Flowwise.
Now let me show you the</b> <b>simplest way to execute all</b> <b>of these previous steps in one single go. </b> <b>Go over to this repository, which I've</b> <b>created specifically</b> <b>for this series, and under branches,</b> <b>select executables. Go to</b> <b>code, and click on download</b> <b>zip.
You can now extract the contents of</b> <b>that zip file anywhere on</b> <b>your machine, and you will be</b> <b>presented with two files - a run. bat</b> <b>file, and a run. shell file.
If you're</b> <b>using Windows, you can</b> <b>double click on run. bat, and for Linux</b> <b>and Mac, you can click on</b> <b>run. shell.
You might see a pop-up</b> <b>like this. Simply click on more info, and</b> <b>then click on run anyway. </b> <b>The first time you run this,</b> <b>it will take a few minutes to complete,</b> <b>because it's installing all the</b> <b>dependencies for the first</b> <b>time, but if you ever want to use</b> <b>Flowwise again, simply double click this</b> <b>file, and the startup</b> <b>will be way faster, and if you want to</b> <b>ensure that you have the latest version</b> <b>of Flowwise installed,</b> <b>all you have to do is delete the Flowwise</b> <b>folder, and then execute run</b> <b>again.
This will not delete</b> <b>any of your existing flows, and will</b> <b>simply download the latest</b> <b>project files and dependencies. </b> <b>The script will automatically open</b> <b>Flowwise in the browser, and now we can</b> <b>get started using Flowwise. </b> <b>Before we create our first Flowwise</b> <b>application, let's have a look at the</b> <b>user interface.
This</b> <b>is the chat flows dashboard, and this is</b> <b>where you can see all your existing</b> <b>Flowwise flows. You are</b> <b>able to search through your existing</b> <b>flows, and you can change</b> <b>your view from this grid view to</b> <b>a list view over here, and we can create</b> <b>a new chat flow by clicking</b> <b>add new, and this will bring up</b> <b>this canvas where we can select from all</b> <b>the available nodes, and</b> <b>we can view settings about</b> <b>this project, and we can click on this</b> <b>chat bubble to interact with our flow,</b> <b>but we will get back to</b> <b>that in a second. Let's go back to our</b> <b>dashboard, and let's have a</b> <b>look at a few other options in</b> <b>this menu.
In marketplace, we can view a</b> <b>list of pre-created</b> <b>templates that we can copy to get</b> <b>started, and under tools we can see a</b> <b>list of pre-created tools</b> <b>that we can plug into our</b> <b>applications. The marketplace is a</b> <b>fantastic way to learn Flowwise, and if</b> <b>we want to use one of</b> <b>these templates, all we have to do is</b> <b>click on the template, and then click on</b> <b>use template. We can</b> <b>now simply save this template, we can</b> <b>give it a name, I'll just</b> <b>call it demo, and then save it.
</b> <b>If we now go back to our dashboard, we</b> <b>can see this demo flow in</b> <b>our list of chat flows. We can</b> <b>delete this chat flow by clicking on it,</b> <b>and then within settings we</b> <b>can click on delete chat flow,</b> <b>and then delete. Let's have a look at</b> <b>tools.
We will go into this</b> <b>in detail during the series,</b> <b>but we are able to assign custom tools to</b> <b>our AI applications, and</b> <b>these tools could be anything,</b> <b>like a tool to fetch the current weather,</b> <b>or the current stock</b> <b>prices. We will have a look at</b> <b>creating these tools in the series. We</b> <b>can use Flowwise to create,</b> <b>manage, and integrate OpenAI</b> <b>assistance, and this is also something we</b> <b>will have a look at in</b> <b>detail in the series.
Let's have</b> <b>a look at credentials. This is where we</b> <b>can manage all the credentials for our</b> <b>project, and as you'll</b> <b>see in the series, we will use several</b> <b>credentials, like the OpenAI API keys,</b> <b>and perhaps connections</b> <b>to other third parties as well, and we</b> <b>can add credentials by</b> <b>clicking on add credential. We can</b> <b>then select the provider that we would</b> <b>like to connect to, and</b> <b>that will store the credential</b> <b>over here, which we can reuse throughout</b> <b>our flows.
We can also use</b> <b>variables to set environment</b> <b>variables and secrets as well. Then</b> <b>lastly, we can create API keys, and this</b> <b>will allow us to call</b> <b>our Flowwise applications using API</b> <b>endpoints. This is also</b> <b>something that we will have a look at</b> <b>during the course of the series.
We are</b> <b>now ready to create our</b> <b>very first AI application using</b> <b>Flowwise.