Introducing AI Native Infrastructure Automation

We need to update the load balancers to be more aggressive with their health checks.

You’ve probably heard this sentence, or one a whole lot like it, many times in your career. What happens after this sentence is complicated, but you can boil it all down to this: a human being translates that simple intent to a series of very complex interactions, using a lot of expert knowledge. The process of actually making that change in a safe, repeatable, automated way remains an obstacle for many of us.

Today, you’ll type that sentence, and System Initiative will simply do it for you. It will discover your load balancers if it doesn’t already know about them. It will map their configuration for you. It will analyze them. Propose a reasonable update to make them more aggressive. Then prompt you to review its work and make the necessary changes.

In the process, it will also have done everything you need to automate them safely over time. Every go-forward change is tracked. Easy reconciliation of drift. Deploy them again and again. Template them. Trigger them from your own workflows. Everything you need to do professional, high-quality infrastructure automation. From one simple, straightforward sentence.

Here’s how it works.

Our custom AI Agent is responsible for taking your prompt, feeding it to the underlying generative AI model, and then using its knowledge of how AWS works, combined with expert knowledge of System Initiative, to make a plan for what needs to be done. It starts by exploring your System Initiative Workspace, searching for relevant infrastructure - like Load Balancers, Listeners, and Target Groups. If you don’t have any, it will dynamically discover them from your AWS account. Then it will analyze the load balancer configuration, determine a more aggressive health check, and propose it to you. You’ll review it to make sure you like the changes, then apply it.

At each stage, the AI Agent is working with our highly detailed, 1:1 models of AWS within a simulated change set. Nothing it proposes can impact the real world without your approval. It uses our understanding of the underlying schemas, validating each configuration and applying any custom policies you have, to make sure the configuration is correct. When the LLM gets it wrong, we detect it and can inform the agent of its mistake, allowing it to course-correct mid-stream. When it comes time to apply the change, you know exactly what is going to happen.

We’ve used it for all kinds of infrastructure automation tasks. Fixing production outages. Automating ticket-based cloud infrastructure helpdesks. Translating a deployment from running on EC2 Instances with Docker containers to using Amazon’s ECS service or Kubernetes (this is probably my single favorite thing it has done - to figure it out, it reads our base64 encoded cloud-init configuration to find the systemd unit file that runs the containers). Suggesting an architecture for you based on an analysis of your source code and documentation. Every day we’re finding new and exciting ways to work.

The AI Agent brings the intelligence, we bring the infrastructure model, and you bring your expertise. Together we’ll build amazing things. Try it for yourself. Then come join us on Discord and share the experience. We’ll be waiting for you.