Starting from a simple Linux console, I configured a Python system connected to Ollama (local LLM), leveraging the Llama3:latest model. With a targeted prompt, the system generates a Terraform script ready to set up a complete environment.
I asked the system to generate what was needed to build an enterprise CRM environment on public cloud, with PHP workers, a relational database, and shell scripts to configure and run the environment automatically.
The process: the AI model generates the main.tf, Python writes to file, I show the GPUs working during generation, final output complete and ready for Terraform.
Once the script is obtained, integration into CI/CD pipelines or shell scripts becomes immediate.
Automation reaches a new level: saves time, reduces errors, ensures high standards for infrastructure configurations. No project data leaves the local machine.