I'll be honest: I was never a big fan of AI-assisted tools. Not for coding, not for daily work. That skepticism changed completely after AWS re:Invent 2025, where I attended sessions and workshops that showcased how tools like Kiro and AWS Transform were reshaping the way cloud professionals work.

Since then, I haven't stopped using them. For several months now, Kiro (both the IDE and the CLI) has become an essential part of my daily workflow.

Kiro is now generally available. It was first launched in preview in July 2025 and reached GA in November 2025, including CLI support. AWS Transform became generally available in May 2025.

What I Actually Do Every Day

As a Cloud Architect, my role is varied. I'm involved in product delivery, customer-facing meetings, and interpreting client requirements. Getting those requirements right is a critical part of my job.

This is where Kiro CLI has been a game changer. In our day-to-day, we constantly run queries, look up resources, and cross-reference information across a client's infrastructure. The AWS CLI on its own can be complex when you need to search and correlate data. Tasks that used to take me two to three hours now take minutes, thanks to Kiro's deep contextual knowledge of AWS services.

That's a real productivity boost, one that lets me focus on what my clients actually need: improving their infrastructure and delivering value to their own customers.

Kiro CLI in Action: Real Examples

To give you a sense of how this works in practice, here are two scenarios from my actual workflow.

Example 1: Finding Untagged EC2 Instances Across Regions

A client asked me to identify all EC2 instances missing the CostCenter tag across 5 AWS regions. With the AWS CLI alone, this would mean writing a multi-region loop, parsing JSON output, and cross-referencing tags manually. With Kiro CLI, I simply typed:

Find EC2 instances without the CostCenter tag across these regions: us-east-1, us-west-2, eu-west-1, eu-central-1, and sa-east-1. Include instance ID, region, and the Name tag in the results.

Kiro understood the intent, ran the necessary API calls, and returned a clean table in under a minute. What would have taken me 30–45 minutes of scripting was done in seconds.

3 instances found missing the CostCenter tag.

Example 2: Generating a Migration Assessment with AWS Transform

During a recent migration project, I needed to assess a client's on-premises VMware environment with over 200 virtual machines. From within Kiro, I invoked the AWS Transform integration and asked:

Run a migration assessment for the imported server inventory. Analyze interdependencies, recommend migration strategies for each workload, and group them into migration waves.

AWS Transform processed the inventory, identified database dependencies, recommended replatforming PostgreSQL workloads to Amazon Aurora, and grouped the servers into four migration waves, all in about two hours. Based on our team's experience, the same assessment would have taken roughly three weeks manually.

Kiro IDE and CLI: A Familiar, Extensible Experience

Kiro IDE runs on macOS, Windows, and Linux. Its interface is built on VS Code, which means the adoption curve for developers is practically zero. I imported my existing settings, themes, and extensions on day one.

What makes Kiro stand out beyond the editor is its extensibility through two key features:

• Model Context Protocol (MCP) servers - an open protocol that lets Kiro connect to external tools and data sources, extending its capabilities far beyond code completion.
• Kiro powers - curated, pre-packaged bundles of MCP servers, steering files, and hooks validated by AWS and partners. Powers give Kiro instant, specialized expertise for specific workflows like serverless development, observability, or infrastructure reviews.

On the CLI side, Kiro CLI brings the same AI capabilities to your terminal. I use it daily for querying AWS resources, troubleshooting infrastructure, and automating repetitive tasks...all through natural language.

AWS Transform: From Weeks to Days

One of the biggest advantages I've found with Kiro is its integration with AWS Transform, a service that uses specialized AI agents to automate complex migration and modernization tasks for VMware, mainframe, and .NET workloads.

A significant part of my role involves cloud migrations - from on-premises environments, VMware setups, or even other cloud providers. The first step in any migration project is the assessment phase: discovering resources, analyzing interdependencies, identifying migration waves, and mapping out which databases and applications move together.

In our experience, a migration assessment that traditionally took our team three to four weeks now takes one to two days with AWS Transform. It handles roughly 80% of what a Cloud Architect would typically do during an assessment:

• Resource identification and inventory
• TCO analysis of virtual machines
• Migration strategy recommendations for database engines (including open-source alternatives like Amazon Aurora)

Being able to invoke AWS Transform directly from Kiro, without switching contexts, is what makes this workflow truly seamless.

AWS Well-Architected Reviews in Under an Hour

Another use case that has been completely transformed: AWS Well-Architected Framework Reviews.

Traditionally, analyzing a client's workload, evaluating it against the Well-Architected pillars, cross-referencing findings, and generating reports took our team one to two weeks. Now, a Kiro power does it in under an hour.

We even built an internal Kiro power that takes the output from the review and generates a PowerPoint presentation using our company's template...complete with critical findings, a supporting PDF document with all the evidence, and actionable recommendations.

A process that took two weeks now takes two days at most. More importantly, it lets us reach clients faster, especially when it comes to:

• Resolving security issues
• Improving workload high availability
• Proposing cost optimization strategies so clients can redirect budget toward delivering value to their own customers

The diagram below illustrates the before-and-after workflow:

Security and Governance Built In

Since Kiro is an AWS service, it falls under the AWS security umbrella. This was a key factor for me and my clients. We can:

• Log all prompts for audit and compliance
• Track token consumption across the team
• Build custom Kiro powers and specs tailored to our company's ecosystem

This means we can standardize how we work with Terraform, specific frameworks, or any tooling a particular client requires, all while maintaining governance and compliance. For organizations with strict security requirements, Kiro integrates with IAM Identity Center for centralized access management.

How I Got Started with Kiro

Setting up Kiro was straightforward. Here's how I did it and how you can too:

Prerequisites

• An AWS Builder ID (free) - or you can sign in with Google or GitHub
• Optionally, the AWS CLI configured with your account credentials - this lets Kiro access and query your actual AWS resources

Installing Kiro IDE

• I downloaded Kiro from kiro.dev - it's available for macOS, Windows, and Linux.
• Installed it like any desktop application
• Signed in with my Builder ID

Since the interface is based on VS Code, I imported my existing settings and extensions. The transition was seamless.

Installing Kiro CLI

On my Mac, I ran a single command:

curl -fsSL https://cli.kiro.dev/install | bashCopyCopied

You can also use Homebrew:

brew install kiro-cliCopyCopied

I verified the installation and started my first session:

kiro-cli version 
    kiro-cli chatCopyCopied

Personalizing My Context

One of the first things I did was set up a context file in my home directory under .kiro/context.md, so Kiro would tailor its responses to my environment. The file contents read:

echo "I am a Cloud Architect working primarily with AWS. I focus on migrations, 
Well-Architected reviews, and infrastructure optimization. I use Terraform for
IaC." > ~/.kiro/context.mdCopyCopied

I loaded it in my session with the /context add command pointed at that file, and Kiro immediately adapted its answers to my role.

Choosing a Model

Kiro gives you access to multiple foundation models. I use the /model command inside a chat session to switch between them - Claude Sonnet 4.5, Claude Sonnet 4, Claude Haiku 4.5, or Auto mode, which picks the best model per task. Each model has different strengths, so I pick based on the complexity of what I'm working on.

From Skeptic to Advocate

I'm no longer skeptical. Kiro and AWS Transform have fundamentally changed how I work. The combination of a familiar IDE, powerful CLI, extensible powers, and deep AWS integration means there's a tool for every part of my workflow.

I now look forward to every new release, knowing each one brings more opportunities to increase my productivity and deliver more value to my clients.

Ready to try it yourself? Start here:

• Download Kiro IDE: kiro.dev
• Install Kiro CLI: cli.kiro.dev/install
• Explore Kiro powers: kiro.dev/docs/powers
• Learn about AWS Transform: aws.amazon.com/transform
• Read the AWS Well-Architected Framework: aws.amazon.com/architecture/well-architected

Have questions or want to share your own experience? Connect with the developer community at community.aws.

Juan Emilio Alemis is a Cloud Architect at Uturn Data Solutions, specializing in cloud migrations, Well-Architected reviews, and AWS infrastructure optimization.