Deploy Full Infrastructure Through a Conversation: No YAML, No Configs

You need a production web server with a database, a load balancer, proper security groups, and auto scaling. You know exactly what you want. Now go build it.

In the console, that's an hour of clicking. Pick an AMI, configure the instance, set up security groups, create the RDS, configure subnets, set up the ALB, register targets, configure health checks, create the auto scaling group. Dozens of screens, hundreds of fields, one wrong setting and you're debugging for another hour.

Or you write config files. Hundreds of lines of YAML or HCL, syntax you'll forget next week. Works, if you have the time and patience to get every bracket right.

Both paths get you there eventually. Neither respects your time.

What if you just described what you want?

"Deploy a production web server running Node.js with a PostgreSQL database, behind a load balancer, with auto scaling from 2 to 6 instances." That's a sentence. That's also a complete infrastructure deployment in Liberra.

Here's what happens: Liberra looks at your existing infrastructure first. Checks your VPCs, subnets, existing security groups, so it doesn't create duplicates or conflict with what you already have. Then it builds a step-by-step plan: every resource it'll create, every configuration, every dependency.

You see the full plan before anything happens. You can ask questions like "why did you pick that instance type?" or "can you use my existing VPC?" and the plan adjusts. When you're happy with it, you approve, and it executes each step in order.

Real provisioning, not a toy

This isn't a wrapper that runs a few CLI commands. Under the hood, Liberra uses a real provisioning engine, the same kind that powers production deployments at scale. It understands resource dependencies, handles creation order, knows that a security group needs to exist before an instance can reference it.

The difference is you don't write the config. You describe the outcome. Liberra translates your intent into the right resources, the right configurations, the right order of operations. And because the AI already knows your infrastructure, it makes smart defaults: using your existing VPC, matching your naming conventions, placing resources in the right subnets.

The point isn't to replace knowing your infrastructure

Liberra doesn't remove you from the process. You still decide what gets built, review every plan, approve every action. What it removes is the busy work: the YAML, the clicking, the "which page was that setting on?" friction.

You know what you want to build. Liberra just gets it done faster.

— Founder, LiberraAI