Opensource AWS alternative cloud platform that is Agent native

[00:00:00]Hello, everyone.

[00:00:01]My name's Abhishek, and welcome back to my channel.

[00:00:05]In today's video, I'll introduce you to an open source cloud platform that is completely AI native.

[00:00:13]Yes.

[00:00:14]Let me explain.

[00:00:16]So, basically, we have seen three generations of cloud computing.

[00:00:20]The first generation, generation of physical servers, where system administrators took care of a lot of manual configuration for us.

[00:00:29]It can be setting up hypervisors, creating virtual machines, storage, networking, everything manually.

[00:00:36]Then we moved to the second generation, the generation of public cloud platforms, where platforms like AWS, Azure, GCP made things little easy for us.

[00:00:47]Here, DevOps or cloud engineers automated these cloud platforms, maybe using scripts like Terraform.

[00:00:53]But now we are moving to the third generation, the generation of AI agents.

[00:00:59]Today, we don't build applications manually. We use AI agents to create the applications. In the same way, we want AI agents to configure the cloud platform for us without any human intervention.

[00:01:12]This is exactly where I'm going to introduce you to open source cloud platform that is AI native in nature.

[00:01:20]I'm talking about InSpoch.

[00:01:23]So, this is going to be a very exciting video. I'll show you how to set up InSpoch and how to configure your AI agent to create resources on the cloud platform.

[00:01:33]Perfect.

[00:01:34]All the useful links for this video are in the description and also in the pinned comment.

[00:01:39]Let's get started.

[00:01:41]So, this is a platform, InSpoch, the AI native cloud infrastructure platform.

[00:01:47]To get started, you just have to follow four simple steps. Step one, click on login, create an account for your personal usage or for your organization.

[00:01:58]The best thing, you can get started with a free account.

[00:02:01]Just use your existing Google account.

[00:02:04]I'll use one of my accounts. And there you go. We have an account set up with In Scribe.

[00:02:10]Once you are here, start with creating a project, which is your step two.

[00:02:15]Now, why are we creating this project?

[00:02:17]Within this project, we will create the resources for our back end or for our application.

[00:02:23]Let me call this as In Scribe demo demo project and choose a region that is nearest to me.

[00:02:31]I'll go with US East.

[00:02:33]Create project. Of course, we are not going to do anything manually, but just to show you, within this project, you can set up authentication, you can set up database for your application, storage, functions, model gateway, any computer related things. Typically, what you get from a cloud platform. That's why it's an agent native open source alternative for AWS.

[00:02:56]Cool. Now, let's head to step three.

[00:02:59]We don't want to do any configuration manually, so we will set up In Scribe CLI and we will link the CLI to this project. So, we will use this NPX command. I'll go back to my terminal and I'll just paste this here.

[00:03:14]Don't worry. This will also set up the In Scribe related packages for you. So, you can see, need to install the following packages. I'll just say yes.

[00:03:23]It is asking for authentication. Let's authorize In Scribe CLI.

[00:03:29]So, it authenticated with my username.

[00:03:31]It linked to the project In Scribe demo project and it also added the required agent skills. See how easy is the configuration.

[00:03:40]So, finally, as step four, I'll go with my AI agent. As I told you, you can use any agent. It can be Claude, GitHub Copilot, Cursor. I'll use Claude. So, I'll just open Claude.

[00:03:53]And I'll just paste this prompt.

[00:03:55]I'm using InScribe as my backend platform. Read the current directory.

[00:03:59]Make sure InScribe skills are installed and use InScribe CLI for the backend tasks.

[00:04:05]This will just verify if Claude is ready with the InScribe configuration. Like, when we create an application with Claude, will it be able to set up our backend on the InScribe platform?

[00:04:17]And you can see here, it says all set.

[00:04:20]Environment is ready.

[00:04:21]Four skills are installed. InScribe, InScribe CLI, InScribe debug, and InScribe integrations.

[00:04:28]These four skills are good enough for you to set up the configuration on the InScribe platform. So, now we are Claude is ready with InScribe as the backend.

[00:04:37]And yeah, let me tell you something important. You don't have to worry about the four simple steps that I followed.

[00:04:43]In fact, I followed these steps from their documentation. So, if you head to InScribe documentation, you'll find the link in the description. You can find the CLI setup where they have all the steps mentioned here.

[00:04:57]Create an account with InScribe. Create the project. Link the CLI. The same command that I've used. Verify the installation where you will run this prompt that I executed on the Claude code terminal.

[00:05:10]So, you can also follow the same instructions.

[00:05:13]Okay. Now that we are done with the initial configuration, let's move to the demo part.

[00:05:19]So, I'll use Claude.

[00:05:21]I'll ask this agent to create a full stack application.

[00:05:24]Let's see how this agent, without any human intervention, will create the required resources for the application on the InScribe platform.

[00:05:32]See, basically Claude can create full stack application. But to make the application production ready, your application would need a database, authentication, storage, compute. All those resources will be created on the InScribe platform. Let's take a look at that. Okay, now what application I'm going to build? So, I'm going to build a local services marketplace where users can look for carpenters, plumbers, electricians, and book them according to the ratings or according to their services. I'll use this prompt. Build a production grade local services marketplace that connects customers with verified plumbers, electricians, and carpenters. Design modern and scalable applications with secure authentication, local based service discovery, real-time booking, online payments, and live order tracking.

[00:06:24]Include separate dashboards for customers, service providers, and administrators with intuitive workflows.

[00:06:31]Focus on clean architecture, responsive UI, performance, security, and exceptional user experience. The platform should be modular, extensible, and ready for real-time production use.

[00:06:43]If you observe carefully, I did not tell Claude anything related to Inswatch, but how would it make use of Inswatch?

[00:06:51]Because we have already configured the skills. Perfect. Now, let's submit the prompt. It started going through the directory. The directory is currently empty, obviously. It provided with the options, which stack for the build. I'll go with Inswatch backend as a service.

[00:07:08]And then it is asking for the payment gateway. For now, I'll just go with the mock payment gateway. I'll submit the answers. There you go. Claude started building this application for us using the Inswatch backend, mock payments, MVP foundation, and it started loading Inswatch tools first. Now, it started with setting up the database, get credentials, load DB reference, and you can see it is making use of the Inswatch DB. Okay, so I'll say proceed with the database. Looks like the initial database configuration is done. Now, it is executing the DB migrations. And in the meanwhile, you can also see how it is creating this database configuration or tables. Like, if you go to the user interface, just refresh, and you can see within the database there are seven tables.

[00:07:58]Heading to the database, these are the tables that are being created. The services table, reviews, provider profiles, payments, bookings, booking events. Right now, there is no data into it because it just started creating the DB configuration and DB migrations.

[00:08:13]And looks like now it is also creating the storage for the application.

[00:08:17]If we go back and refresh the dashboard, so I can see one storage bucket here.

[00:08:24]Yeah, one bucket.

[00:08:26]So, till now, it created the database configuration and also the storage configuration. And as it started working more on the front end, we can see uh now we are at a point where the agent wants to implement authentication for the application. So, it has three options for us. We definitely want to set up the authentication. So, I'll go with the second option, keep authentication, where it is going to set up the email authentication login workflow. And you can see the moment I opted for secure authentication, Cloud started thinking how to handle email verification for the demo. Keep verification SMTP. And it started looking for this skill in swatch auth SDK integration.md.

[00:09:12]So, probably now this will go ahead and create authentication on the back end.

[00:09:18]Yeah, as expected, so it is using the In Swash library. While it is implementing the authentication, one thing that I wanted to mention, so if you go back to the platform, and let's say you want to add some sample data to this database or to the storage, you can just copy this prompt, use In Swash skills to create a table in in forge back end with some sample data.

[00:09:42]You can just copy it copy this prompt and you can paste it to your agent and it would also populate your application with the sample data maybe for your MVP or demo purpose. Anyways, going back.

[00:09:56]So, it says I can read all.users. Plan create demo authentication users via sign up and seed public rows keyed by the email.

[00:10:06]Fine. Now the authentication is also set up. So, if you go back and if you head to the authentication, so you can see it created some sample users and it enabled email authentication. So, anyone can authenticate to this web application that we're going to create using their user ID and it created some sample user IDs just for our demo purpose. So, we are done with authentication, we are done with the database setup and also the front end setup. Now we are only left with some basic flows like the customer flow, search, booking, payment, tracking.

[00:10:42]Probably we should be done in next 2 to 3 minutes of time. Until now, everything is very flawless.

[00:10:48]There you go.

[00:10:50]So, we have the complete application ready. It says the status as delivered.

[00:10:54]For back end, it used the in forge project typically for database storage and front end it developed using React, but for authentication, again it used the in forge platform.

[00:11:05]Now let's run this application and try to access this application on the browser. So, we will use the NPM command to run the application.

[00:11:13]Okay, now it is running on localhost 5173. Let me try to access this.

[00:11:18]Here we go. So, this is our the local services platform.

[00:11:23]And here you can find plumbers, electricians and carpenters. I'll click on the plumbers. Okay, I can select the location. Let me go with San Jose.

[00:11:33]There are no plumbers in this location.

[00:11:35]If I go with San Francisco, there are some plumbers in the location.

[00:11:38]Okay, I can also use this toggle to find the local services within specific range. I have the authentication workflow here. I can click on sign in.

[00:11:49]I can use one of the existing email addresses or we can just create an account. Just click on create an account, provide your email details, and you will have your account ready. Let's see by providing basic information. So, I'll just click on create an account.

[00:12:05]Perfect. So, it sent a six-digit verification code. So, our authentication workflow is very clean.

[00:12:11]Let's verify and see. Great. Now, I'm logged in as a user, and we should see that entry within the database. Here it is. Abhishek, the new user, is added and stored to the users table. So, the authentication workflow is great.

[00:12:25]Database's workflow is also great because we have some plumbers and electricians populated here. We can see a lot of information on the dashboard.

[00:12:34]All of this is coming from the database.

[00:12:37]So, this is cool. In just 15 minutes, we not only created the application, but we created a production-ready application with typical production flows like authentication, storage, and database.

[00:12:49]And interestingly, let's say if you want to add an AI assistant, Insoch also provides a model gateway.

[00:12:56]So, you can just pick up your model. It supports OpenRouter as well. So, by default, you get this $1 credit so that you can perform a proof of concept.

[00:13:07]You can make use of it. So, using this, you can also set up AI assistant on your platform, something like a cool assistant where users can come and ask questions to it.

[00:13:18]And finally, if you want to deploy this application, let's say you want to make it live to the users.

[00:13:24]So, you You have to tell Insoch, "Deploy this application to the end users."

[00:13:30]And it will also take care of the deployment and bringing your application live.

[00:13:35]Let's see.

[00:13:36]So, the moment I provided this instruction, again, Plot figured out a skill. So, within the Insooch CLI, within the deployments, it could find the deployment skill, and it started the deployment process as well.

[00:13:49]If you're interested in understanding under the hood, it is basically making use of Vercel.

[00:13:53]And what makes it easy? You don't have to configure anything within Vercel.

[00:13:58]You can see here, just after 30 seconds, it says the deployment status is ready, and this is a URL. Now, instead of localhost, I'll just copy this URL, and I'll access this URL from the browser.

[00:14:12]The same application. Isn't that cool?

[00:14:15]Definitely, yes.

[00:14:17]Finally, just a quick summary of what we learned in the video.

[00:14:21]So, we learned about Insooch, an agent-native, open-source cloud platform.

[00:14:26]It's very easy to get started. You just have to create a free account.

[00:14:30]Just create a project as we did in the video.

[00:14:33]Then, download the Insooch CLI and connect any coding agent that you're using. It can be Claude, Cursor, Copilot, Codex.

[00:14:42]You can just connect with your Insooch project, and the agents will start using Insooch as your backend. And this way, your agent can not only create an application, but it can create production-ready application with features like authentication, database, storage, and model gateway, or AI assistant for your website. So, I highly encourage everyone to try out Insooch.

[00:15:06]Once you're comfortable with it, let's say for your demo application or for your proof of concept, you can go ahead, create a new organization, and you can upgrade your plan on Insooch. If you have any questions about today's video, do let Let in the comment section. I'm more than happy to help. All the useful links for the video are in the description and also in the pin comment.

[00:15:28]See you all in the next video. Take care.