Autonomous Infrastructure: Deploying LLMs in the Cloud

This project was developed for the IFTech 2026 presentation. Its goal is to demonstrate how to provision a scalable cloud infrastructure using Infrastructure as Code (IaC) and deploy a private Large Language Model (LLM) integrated with an interactive application for participants.

The repository is structured as a Monorepo, combining both the infrastructure layer (IaC) and the application layer.

---

Repository Structure

iftech/
├── iac/                       # Infrastructure as Code
│   ├── terraform/             # Cloud resource provisioning (VM, VPC, Security Groups)
│   └── ansible/               # Server configuration (Docker installation and dependencies)
└── app/                       # Application and Artificial Intelligence
    ├── frontend/              # Web interface for participant interaction
    ├── backend/               # API integration with the LLM
    └── docker-compose.yml     # Local service orchestration (App + Ollama/vLLM)

Technologies Used

• Terraform: Virtual machine provisioning and network rule management.
• Ansible: Server configuration automation and Docker provisioning.
• Docker & Docker Compose: Containerization of the application and the LLM inference environment.
• Ollama / vLLM: Local inference server for running the LLM in the cloud.
• Python (FastAPI / Langflow): Backend and AI agent orchestration.

Authors

• Ryan Morais Correia – Infrastructure Lead (Terraform & Ansible)
• João Pedro Marques – Application & AI Engineering Lead