UniqureAI

UniqureAI is a web application providing highly realistic, low-latency AI companions. With stunning dynamic designs, a 3D visualizer, and sub-second voice interactions, it creates a fully immersive AI connection.

Live Application: https://uniqureai.web.app

Features

• Real-Time Voice AI: Ultra-low latency voice calling with Speech-to-Text (Whisper via Groq), LLM processing (Llama 3 via Groq), and Text-to-Speech (Cartesia) streamed over Server-Sent Events (SSE).
• Interactive 3D Visualizer: A custom WebGL Three.js visualizer that responds in real-time to both the user's microphone frequency and the remote AI's voice frequency.
• Companion Categories: Filter dynamic AI Personas by "Love of Life", "BFF", and "Mentor", with fully implemented gender preference matchmaking.
• Dynamic Swipe Feed: A "Discover" feed to match with AI Personas.
• Mini-Games: Play real-time games with your AI companions like Would You Rather, 20 Questions, Escape Room, and Story Builder.
• Premium UI/UX: Dark-mode aesthetic with neon green (#CCFF00) accents, fluid page transitions, glassmorphism, and micro-animations built with Framer Motion and Tailwind CSS.
• Progressive Web App (PWA): Installable to the iOS/Android Home Screen for a completely native, full-screen hardware-accelerated experience.

Technology Stack

• Frontend: Next.js 14 (App Router), React, Tailwind CSS, Framer Motion, Zustand
• 3D Graphics: Three.js, React Three Fiber, React Three Drei
• Real-Time Comm: LiveKit (for STT/TTS audio streaming)
• Backend: Node.js, Express, Firebase Functions (Cloud Run), Firebase Hosting
• Database/Auth: Firebase Firestore, Firebase Authentication
• AI Models: Groq (Llama 3, Whisper), Cartesia (Sonic TTS)

---

🚀 Setup & Installation

This repository contains both the frontend (Next.js) and the backend (Express) services. You will need to run both concurrently for full functionality.

1. Prerequisites

• Node.js (v20 or higher recommended)
• A Firebase Project (with Firestore and Authentication enabled)
• API Keys for Groq and Cartesia

2. Environment Variables

For security, the actual API keys have been removed from this repository. You must create your own .env files in both the frontend and backend folders.

Frontend (frontend/.env.local):

NEXT_PUBLIC_FIREBASE_API_KEY="your_firebase_api_key"
NEXT_PUBLIC_FIREBASE_AUTH_DOMAIN="your_firebase_auth_domain"
NEXT_PUBLIC_FIREBASE_PROJECT_ID="your_firebase_project_id"
NEXT_PUBLIC_FIREBASE_STORAGE_BUCKET="your_firebase_storage_bucket"
NEXT_PUBLIC_FIREBASE_MESSAGING_SENDER_ID="your_firebase_sender_id"
NEXT_PUBLIC_FIREBASE_APP_ID="your_firebase_app_id"
NEXT_PUBLIC_BACKEND_URL="http://localhost:8080"

Backend (backend/.env):

CARTESIA_API_KEY="your_cartesia_api_key"
GROQ_API_KEY="your_groq_api_key"
PORT=8080

3. Running the Backend (AI API)

The backend handles the secure server-sent events for AI chatting and voice generation.

cd backend
npm install
npm run dev

The backend will start running on http://localhost:8080.

4. Running the Frontend (Next.js App)

The frontend contains the UI and connects to the backend.

cd frontend
npm install
npm run dev

The frontend will start running on http://localhost:3000.

5. Deployment

This project uses Firebase Hosting for the frontend (which automatically bundles the Next.js App Router into a serverless Cloud Function) and Google Cloud Run for the standalone backend.

# To build and deploy the Next.js frontend to Firebase Hosting
cd frontend
npm run build
npx -y firebase-tools@latest deploy --only hosting

---

📸 Screenshots

Authentication & Profile

Discovery Categories

Real-Time AI Chat

Liked You & Mini-Games