Transcription
Around a month ago, we had taken a look at Sim AI, another platform for easily creating AI agents that can automate anything. In a way, it's like an open-source Nathan alternative that's simpler to use, very similar to Nathan, but not as technical.
But some people are looking for a technical and easy to use balance. Which is why today I'm back to showcase a new alternative to both Nathan and Simei. A workflow automation and orchestration tool that's open- source, fully local, and seriously powerful. It's called Kestra, a fast, scalable, declarative AI native orchestration platform where you can practically automate anything.
You may be wondering what makes Kestra different from something like Nathan. Well, unlike Nathan which is an integration first with a nodebased UI, Castra is orchestration first which is bringing infrastructure as a code to data pipelines infra as well as event-driven workflows so you can build reliable automations in a few lines of VM right from the UI. And you can even have AI take on this task of doing it for you which is pretty easy and simple.
For example, let's say you got a load of CSV files that would normally take hours in Google Sheet to orchestrate, configure, as well as calculate different sorts of formulas. But with Kestre's no code AI copot, you can give it a prompt as to what you want to accomplish in plain natural language. And it builds the automation for you, handling all the calculations automatically, the formatting with a new file. And it will also even tell you when a new file has been created by sending you messages through different platforms if you require like Slack messages when the workflow is actually completed.
And remember guys, Kestra is open-source and fully available locally. It's a platform that blends code flexibility with a noode UI and it's super easy to get started. You have 950 plus integrations, a lot of different services and APIs available. Uh you can also configure a lot of things on your own which is pretty helpful and useful. But to get started it is super simple. All you need to have installed beforehand is Docker Desktop as a prerequisite. You can install it for whatever operating system you may have, but if you are going to be using the Docker method, you can simply follow through with this quick start method, which I'll leave a link to in the description below.
There's also a cloud hosted option, but it's paid, I believe, and I don't really recommend you use that cuz you if you have the ability to use it locally, much better that way cuz you can access it for free. So, to get started, just make sure you have Docker Desktop installed. It's completely free to install, which I'll leave a link to in the description below.
Once you have it installed, if you are using Docker and if you are looking to get started with it, you can easily just copy the script over here, which is a single command that you can run within your terminal. But there's a couple of different presets that you can use if you're on Windows. So, if you're using PowerShell, you can use this command set. If you're on Windows and you're using command prompt, you can use this. And if you are on Windows and you're using WSL, you can copy this, which is something that I'm doing. So, I'm going to copy this. Make sure I have Docker Desktop running on my toolbar, you can check that by opening this up and see if Docker Desktop is running, which it is. And now what we can do is simply open up WSL. And then now we can paste in this command, which is going to work on installing Castra locally on our computer. It's easy as that, guys. Once it finishes pulling and installing all the containers, we can then open it up within our local host.
Before we get started, I just want to mention that you should definitely go ahead and subscribe to the world of AI newsletter. I'm constantly posting different newsletters on a weekly basis. So, this is where you can easily get up-to-date knowledge about what is happening in the AI space. So, definitely go ahead and subscribe as this is completely for free. There we go. We have catcher running locally.
Now, what you got to do first is just simply create an account so that you can instantly access Castra locally with your own authentication. So, simply go ahead and fill this out. After creating an account, this is where you're going to be sent over to the main dashboard of Kestra where you can now get started and create your automation flows.
Now, what you want to do first is take a look at all the different panels. On the main dashboard is where you can access all your workflows. You can access all your flows over here. These are a couple of tutorial flows that you can actually readily access and gives you a better idea of what is happening. You can execute this AI agent calling flow by simply starting it up, but we're not going to be doing that just yet. You have all your apps where you can build custom apps with Castra, which is great. You have executions where you can manage all of them. Uh see the total runtime as well as when it was logged. You have a couple of tests that you can visualize. name spaces, the KV store which is the database, secrets where you can manage all all of your environment keys. You have blueprints which are templates in a way that you can readily access uh whether that's running a specific task uh triggering something with different automations and you can even manage different plugins with 713 plus plugins which is insane.
But now on the main dashboard, if you click on create your own workflow from scratch, you can use the tutorial to help you build anything. This is where it's going to teach you what you can do, all the use cases like where you can build AI agents, business operations like automation, sorry, uh business process, data engineering, etc. We're going to go ahead and create a simple AI agent. We're going to click next. And this is essentially the code editor. Now, you don't need to write any bit of code because the AI copilot is something that you can use. You're just going to need to configure it within the application yam file by setting your Gemini or whatever provider's API key and then you can simply use it to build almost anything. But this is where you can visualize the different nodes of your automation on the right hand section. Now, there's also a playground that you can enable, but you can also execute it by clicking on this button, and then you can launch the AI agent that you're trying to create. Now, it's going to fail because I haven't set my API key. But then you can start creating your new flows and start building automations exactly from here, but just take a look at this flow that was already created, which is the business automation tutorial. And essentially this is where it's able to process anything from your working directory going to be able to query it and it's going to process your CSV file and essentially it's converting the CSV data to the Google sheet that we had showcased at the start. Its purpose is to automate repetitive business tasks like data processing, notification, approval. And essentially once you have set the different configurations within the execution and edit flow, you're going to be able to then have it so that you can execute this flow which is going to automate this process. So if you click on execute, it's going to work on processing the working directory CSV files. It's going to then query their prompt which is to create a table in this case or delete certain features a part of the CSV file and then from CSV it will proc it to the Excel Google sheet.
But when you're starting off from scratch if you're working on the node code section where you're configuring each of the nodes just simply define the ID of the node as well as the description as to what you're trying to achieve. But within the flow code this is where you can configure the YAM file with AI copilot. This is essentially where you can just naturally describe what you want to configure and the AI will basically take care of the task of handling, what sort of automation you want to build, code it out for you and then you can visualize it over here and then try it out within the playground which you can then run as you execute this task.
Here is a data engineering pipeline. Essentially, it is taking in dummy data and from here it is processing it by transforming the data into a specific pipeline. But what's happening here is where Castra is actually fetching the JSON data set from the remote API and its purpose is to basically simulate ingesting raw data from external sources like this dummy JSON website and from here it is then transforming the data which Kestra does autonomously with the AI agents. It runs a Python script inside this temporary docker container and its purpose is to isolate as well as make sure that the environment is consistent and fully automated by using these different debuggers. And then lastly it is the query which is where it is focusing on running this SQL like query against the transformed data that we had sent in and its purpose is to automate the analytics after ingesting and transforming that data. And all of this was fully controlled with the help of Castra.
If you like this video and would love to support the channel, you can consider donating to my channel through the super thanks option below. Or you can consider joining our private Discord where you can access multiple subscriptions to different AI tools for free on a monthly basis, plus daily AI news and exclusive content, plus a lot more. But that was just a simple overview of what Kestre is capable of doing. There's so much more to this, so I highly recommend that you take a look at it with the links in the description below. This is something that's open source, fully local, and it's a bit more technical, but it gives you a lot more flexibility as to what you're trying to achieve with Kestra. So, make sure you take a look at it with the links in the description below. Subscribe to the second channel, join the newsletter, make sure you subscribe to the Patreon page to access our Discord, follow me on Twitter, and lastly, make sure you guys subscribe to the YouTube channel. Make sure to turn on notification bell, like this video, and please take a look at our previous videos cuz there's a lot of videos that you will truly benefit from. But with that thought, guys, thank you guys so much for watching. Have an amazing day. Spread positivity and I'll see you guys fairly shortly.