Reimagining Cloud Architecture with GenAI

Editor’s Note: This blog is adapted from a talk by Amoghavarsh Patil, a cloud architect and creator of Cloud Sketcher. In this session, he shared how he built an AI-powered tool to simplify and accelerate cloud architecture design across Azure, AWS, and GCP. From sketching diagrams to generating Terraform code and real-time cost estimates, Cloud Sketcher bridges GenAI with practical infrastructure work, reshaping how architects think, plan, and build.

Hello. I am Amoghavarsh Patil, a cloud architect, engineer, and someone who enjoys building tools in my spare time. Over the last two years, I have been working on a side project that has grown into something larger than I expected. It is called Cloud Sketcher, and it is built to make cloud architecture simpler, faster, and more intelligent, especially for those working across multiple cloud providers.

This is not just a tool. It is my answer to a very real and persistent pain in cloud consulting: the time it takes to design, translate, and document architecture across Azure, AWS, and GCP.

Cloud architects are often expected to deliver diagrams and deployment strategies for all three major cloud providers. A client might say, "Here is the infrastructure I need. Show me how it looks on Azure, AWS, and GCP." They usually only choose one, but they want to compare all three first, in terms of structure, cost, and complexity.

This process used to take me hours, sometimes days. I would redraw the same system three times, switch between tools like Draw.io or Lucidchart, and recheck every resource mapping by hand.

That is when the idea hit me. What if I could draw it once, and let the tool translate it? What if I could go even further—and generate the Terraform code, the documentation, and even cost estimations from the same sketch?

I started building Cloud Sketcher in August 2023. At the time, GenAI tools like ChatGPT were just starting to gain momentum, and I knew they had the potential to change the way we approach infrastructure design.

Cloud Sketcher is the first AI-powered cloud architecture tool built specifically for Azure, AWS, and GCP. It is not a generic flowcharting app. It understands cloud patterns. It speaks architecture. It does three key things:

• It lets you sketch cloud diagrams
• It converts the design across cloud providers
• It generates Terraform code and documentation based on that design

One of the tool’s core features is cross-cloud translation. If I draw an Azure-based architecture, I can convert that to AWS or GCP with a single click. Cloud Sketcher handles the resource mapping, adjusts icons, and aligns configurations with provider-specific best practices.

This is not just helpful for visual consistency. It saves an enormous amount of time during pre-sales, internal POCs, or cloud migration planning. For architects unfamiliar with a particular cloud, it offers a fast starting point—one that is more than just a guess.

Four major modules inside Cloud Sketcher make it more than a static tool:

Sketch Draw

Provide a simple problem statement—like "I want to host an e-commerce app"—and the tool generates a draft architecture. It knows to include a CDN, a web app layer, backend services, a database, and essential components like Azure Functions or Log Analytics.

Sketch Vision

If you already have an image-based architecture diagram, upload it. Cloud Sketcher turns it into an editable canvas. You can modify icons, edit connections, and layer on new components—all without starting from scratch.

Sketch Docs

Once your diagram is done, generate a document instantly. It includes technical and non-technical summaries, architecture overviews, and even cost estimates for development or production environments.

Sketch TF (Coming Soon)

This upcoming module will allow users to export their diagram into Terraform configurations. You will be able to select instance sizes, machine specs, and deployment regions before generating a complete, deployable module.

There are many diagramming tools in the market—Draw.io, Lucidchart, and Eraser. But they are generic by design. Cloud Sketcher is purpose-built for cloud engineers. It only supports Azure, AWS, and GCP. It comes with over 1,000 cloud icons, 150+ MAP icons, and is structured for real-world use, not just internal visuals.

Here is what sets it apart:

• Draw once, convert across clouds
• Document your design with one command
• Estimate costs in real-time
• Export production-ready Terraform (in the next release)

This kind of convergence is not about convenience. It is about accelerating decisions, especially when time and accuracy are critical.

Every cloud provider offers more than 300 services. No single architect has full mastery of them all. Cloud Sketcher helps by surfacing the most relevant components for a given use case—whether you are deploying a microservices backend, hosting a static site, or planning a compliance-driven setup.

The tool is not meant to replace deep cloud knowledge. It is designed to get you 70 to 80 percent of the way there—fast. That is where real efficiency comes from.

Since launching three months ago, Cloud Sketcher has seen adoption across more than 2,000 users. Engineers from IBM, Bosch, Aspire, and several startups have begun using it in their design workflows. I recently received backing through the Microsoft for Startups program, and I am continuing to improve the platform week by week.

Whether you are an architect, a DevOps engineer, or a developer just trying to understand the shape of a system, I believe this tool can help you move faster, think clearly, and design better.

Cloud Sketcher is free to use and available at www.cloudsketcher.com.
You can:

• Draw diagrams for any cloud
• Translate across providers
• Generate documents and estimates
• Soon, export Terraform modules

If you want to connect, share feedback, or collaborate, find me on LinkedIn. I would love to hear how you are designing the cloud.