This is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady
What if you could design cloud infrastructure by simply having a conversation? ☁️💬
I've been building something I call Infrastructure as Chat, an experiment exploring how we interact with cloud infrastructure.
Instead of manually writing Terraform configurations, imagine describing your architecture in plain English:
💬 “I need a web application with a load balancer, three application servers, Redis, and a PostgreSQL database.”
The application translates that conversation into structured infrastructure definitions and generates Terraform code.
But the interesting engineering challenges go beyond generating configuration files.
I'm exploring how to make this workflow more useful through:
- 🏗️ Natural language → Infrastructure: Translating requirements into infrastructure definitions.
- ⚙️ AI + Terraform: Generating infrastructure-as-code that developers can review and modify.
- 🔍 Validation & reasoning: Moving toward understanding dependencies, potential bottlenecks, and architectural trade-offs.
- 🚀 A more conversational developer experience: Making infrastructure easier to reason about without hiding the underlying engineering.
The bigger idea is that AI shouldn't just write infrastructure code. It should help developers understand the systems they're building and the consequences of their decisions.
Still a work in progress, but this has been a fascinating intersection of AI agents, backend engineering, cloud infrastructure, and developer tooling.
Building, breaking, and learning. 🛠️
Top comments (0)