Steering to Autonomy

This post summarizes the opening talk at Autocon/23 by Suresh Mathew, who highlighted the transformative impact of autonomous cloud management. He presented autonomy as a solution to ongoing issues in cloud operations, such as resource waste and preventable incidents, areas where traditional scaling and management fall short. Suresh emphasized how autonomous systems, like Sedai enhanced by Large Language Models (LLMs), empower companies to make smarter, safer, and more cost-efficient cloud management decisions. His message was clear: transitioning to autonomy is about giving teams the right tools to enhance reliability, reduce waste, and deliver seamless user experiences. The evolution from early tools like ECS and Kubernetes to VM support across cloud platforms demonstrates the critical role of autonomous infrastructure in the future of cloud operations.If you would like to watch the video please click: https://sedai.wistia.com/medias/fdnv0lnmdo

Let's explore an important statistic that directly impacts our businesses. In 2015, this number was 35. By 2017, it dropped slightly to 32, only to rise again to 33 in 2022. Now, imagine for a moment that this figure represented something as serious as cancer rates or road accidents. We would undoubtedly investigate further, conduct in-depth research, and likely implement systemic changes to address the issue.

So, what is this number? It represents our bottom-line wastage. For years, it has hovered around the 35% mark across our companies. This is a critical figure that we have the power—and the responsibility—to address. By reducing wastage, we can make a substantial difference, not only to our bottom line but to the efficiency and sustainability of our businesses.

How Autonomous Systems Are Changing the Game

Now, I want you to look at another number, a more hopeful one. This company adopted autonomy just a year ago. Look at where they started, and notice something amazing. Their numbers don’t fluctuate anymore. It’s consistently going down. Autonomy is helping them reduce waste—consistently.

Preventing the Preventable

Here’s something else to think about—8 out of 10 incidents are preventable. Yet, we’ve built an entire industry just to handle the aftermath, the post-mortem. Why do we expect incidents to happen instead of preventing them? We focus so much on cleaning up the mess that we don’t ask ourselves the most important question: how do we stop this from happening in the first place?

‍

A Real-Life Example of Reliability

Let me show you what happens when we focus on the right things. This is a public company that achieved an astonishing reduction in incidents—98% to be exact—and their reliability jumped from 99.9% to 99.99%. Think about that for a second. That kind of change doesn’t happen by accident. It’s the result of embracing autonomous systems. And it wasn’t easy for them, but the results are undeniable.

Why We Need the Right Tools for the Job

Here’s the thing—we’ve tried every tool in the book over the last 10 years. Public cloud, cluster managers, Infrastructure as Code (IaC), scaling up and down—name it, we’ve tried it. And yet, we’re still wasting resources. Year after year, we clean up and fine-tune, but something’s missing. The problem isn’t our engineers or our technology. The problem is we’re using the wrong tool for the job.

The Importance of Autonomous Systems

It’s not about how smart we are or how advanced our tech is. The issue is that we need the right tools. And those tools are autonomous systems. We need systems that filter out the noise before it reaches our developers or operators. Without autonomy, you’re setting your team up for failure.

Our Journey with Cloud Platforms

"At Sedai, we started our journey with technologies like ECS, Lambda, and Kubernetes on AWS, Google, and Azure. And today, I’m excited to announce that we’re expanding our support to AWS VMs and Azure VMs. And soon, we’ll be rolling out support for Google VMs, including stateful apps across multiple clouds. This is a huge step forward, allowing us to bring the power of autonomous systems to even more platforms."

FinOps: Moving Beyond the Wall of Shame

Let me talk a little about FinOps. For the last 10 years, FinOps has felt like a 'wall of shame,' right? A long list of things that we haven’t done. It’s a daunting list of action items that just keeps growing. But here’s the truth: with autonomy, you don’t have to keep working through that list forever. You need a system that cuts that list down to something manageable, something your operators can handle.

‍

Introducing Sed with LLM: Smarter, Safer Decisions

Now, let’s talk about something really exciting—Sedai with Large Language Models (LLM). We’ve enabled LLM on Sed, which means you can now interact with Sed in a completely new way. You can ask it questions like, 'Is this change safe to make?' or 'How many Lambdas are taking traffic?' or 'What’s the cost of our Kubernetes clusters?' Sed can be part of your change control board, helping you make smarter, safer decisions. This is a game-changer for any company managing large-scale cloud environments.

Conclusion

At the end of the day, the future of operations isn’t about hiring more engineers or adding more tech. It’s about embracing autonomy. It’s about giving your teams the right tools to do the job efficiently. We’ve seen the results—98% reductions in incidents, dramatic improvements in reliability, and a steady decrease in wastage. The question now is: how fast can you adopt these systems to transform your business?