AeroResolve — Zero-Touch Agentic Travel Recovery

Autonomous AI Agent for Managing Travel Disruption
Built for IIIT Sri City IdeaVerse Hackathon 2026

Team: Code Alchemists

• Vikhyat Gupta
• Gurnoor Singh Bagga
• Mohit Choudhary

Tech Stack

• LLM: Groq (Llama 3.1 8B Instant)
• Orchestration: LangGraph (4-agent pipeline)
• Backend: FastAPI + Python
• Frontend: React + Vite
• Notifications: Twilio SMS
• Flight Data: AviationStack API
• Infra: Docker

Architecture

Flight Status Monitor (AviationStack)
  → Assessment Agent (analyze disruption impact)
    → Policy Agent (validate rebooking rules)
      → Solution Agent (find & rank alternatives)
        → Execution Agent (book + send SMS via Twilio)

Quick Start

1. Setup Environment

# Fill in your API keys in .env
cp .env.example .env  # or edit .env directly

2. Run with Docker

docker compose up --build

3. Run Locally (without Docker)

# Backend
cd backend
pip install -r requirements.txt
uvicorn main:app --reload --port 8000

# Frontend (in another terminal)
cd frontend
npm install
npm run dev

4. Demo

1. Open http://localhost:5173 (React dashboard)
2. Click "Simulate Disruption" on any booking
3. Watch the 4 agents execute in sequence
4. Receive SMS notification on your phone

API Endpoints

Method	Route	Purpose
GET	/	Health check
GET	/bookings	List monitored bookings
POST	/simulate-disruption	Trigger full recovery pipeline
GET	/pipeline-status	Current pipeline state
POST	/start-monitor	Start background flight poller
POST	/stop-monitor	Stop background poller

Contributions

Role	Name
Architecture, implementation & all code	Vikhyat Gupta
Ideation & presentation	Gurnoor Singh Bagga, Mohit Choudhary