meamir/queuestorm-api

QueueStorm API - CRM Ticket Triage

An automated, high-performance ticket triage system built with FastAPI and powered by Google Gemini 2.5 Flash. This application automatically classifies incoming customer support tickets into predefined categories (e.g., Technical Support, Billing, Account Access) and falls back to a precise keyword-matching algorithm if the LLM provider experiences network or credential issues.

🚀 Live Deployments

• —Primary (Vercel Serverless): https://your-vercel-project-url.vercel.app
• —Secondary / Backup (Hugging Face Docker Space): https://meamir-queuestorm-api.hf.space

⚙️ Features

• —AI-Powered Triage: Instant text classification using the lightweight and fast Gemini 2.5 Flash model.
• —Fail-Safe Fallback: Deterministic regex/keyword matching backup system to ensure zero-downtime routing.
• —High Concurrency: Asynchronous request handling via ASGI/Uvicorn.
• —Automated Docs: Self-documenting interactive API playground out of the box.

🛠️ Local Setup Instructions

If you are a judge or organizer testing this application locally, follow these steps to spin up the server in under 2 minutes.

1. Prerequisites

• —Python: Version 3.9 or higher
• —Credentials: A valid Google Gemini API Key

2. Installation & Environment Configuration

Clone this repository, navigate to the folder, create a virtual environment, and install dependencies:

# Clone and enter directory
git clone <your-repository-url>
cd SUSTMock

# Set up virtual environment
python3 -m venv venv
source venv/bin/activate

# Install required packages
pip install -r requirements.txt