Preflight AI

A Dual-Model, AI-Powered Infrastructure Risk & Chaos Engine.

Preflight AI analyzes your Terraform deployments to identify critical failure impacts before they happen. It parses infrastructure as code into a mathematical graph, simulates cascading failures using BFS, and leverages a dual-model AI architecture to assess reliability, security, and cost risks.

✨ Core Features

• Topology Graph Generation: Instantly parse .tf files into mathematical Directed Acyclic Graphs (DAGs) to visualize your exact architecture.
• Dual-Model AI Engine: Uses Fireworks AI (DeepSeek-V4-Pro) for logic synthesis and a self-hosted Llama-3-70B on an AMD MI300x Developer Cloud GPU for high-volume Chaos Engineering Monte Carlo simulations.
• Blast Radius Analysis: Click any node to simulate an outage and instantly see upstream/downstream cascading failures.
• Automated PR Code Review: Prevent bad infrastructure from being merged by adding the Preflight GitHub Action directly to your CI/CD pipeline!

📚 Documentation

Detailed documentation on how the system works and how to set it up:

• Architecture & Dual-Model AI - Learn how we route workloads between Fireworks and the AMD Developer Cloud.
• Deployment Guide - Instructions for deploying the EC2 backend, AMD droplet, and Vercel frontend.
• GitHub Actions Integration - How to add Preflight AI to your own repositories to automate PR reviews.

🛠️ Quick Start (Local Development)

1. Clone this repository.
2. Create a .env file in the project root:

   FIREWORKS_API_KEY=your_key_here
   LLM_API_BASE=http://your-amd-droplet:8001/v1
   LLM_API_KEY=amd-demo-key
   LLM_MODEL_NAME=meta-llama/Meta-Llama-3-70B-Instruct

3. Start the backend: cd backend && pip install -r requirements.txt && uvicorn main:app --reload
4. Start the frontend: cd frontend && npm install && npm run dev
5. Access the web dashboard at http://localhost:5173.

---

Built for the AI Infrastructure Hackathon.