Relive HomePage

🧠 Relive – AI Memory Companion

💡 Inspiration

In a world where memories are fleeting, Relive empowers people to reconnect with voices, moments, and relationships they've lost or wish to preserve. Whether it's speaking to a late loved one, reconnecting with your 10-year-old self, or hearing the encouraging voice of your future self — Relive turns memories into meaningful, emotionally intelligent conversations powered by AI.

We imagined a future where technology doesn’t just assist us, but remembers with us.

💬 What it does

Relive is a powerful, emotionally intelligent AI memory companion that enables users to:

📖 Create AI personas of loved ones, future selves, or custom characters
💬 Have real-time, context-aware conversations powered by Gemini Pro
🧠 Get emotionally intelligent responses matching traits, backstories, and tone
🔊 Hear personas talk through ElevenLabs voice cloning
🎥 (Coming soon) See avatars respond via Tavus video agents
🔐 Store memories, chat logs, and personas securely using Supabase
🧾 Access full features with subscription via RevenueCat

🏗️ How we built it

✨ Frontend

React 18, Vite, TypeScript
TailwindCSS 3.4.1 + Framer Motion for Apple-grade animations
Radix UI, Lucide Icons, React Hot Toast
Mobile-first, glassmorphism-inspired, WCAG-compliant design

🔧 State & UX

Zustand – local + global state
React Query – server sync and caching
LocalStorage – demo mode persistence

🧠 AI & Multimedia

Gemini Pro + Flash – long-term memory + fast replies
ElevenLabs – real-time, cloned voice responses
Tavus – video avatar generation and conversational video interface (CVI)

🗄️ Backend & Auth

Supabase (PostgreSQL) – RLS, auth, realtime DB, auto-APIs
Row Level Security for user isolation
Email/password login, persistent session tokens

💸 Payments & PWA

RevenueCat SDK – paywall & subscription logic
PWA-ready: installable, offline support, mobile gestures
Netlify: one-click full-stack deployment

🚧 Challenges we ran into

Designing emotion-aware conversations that reflect each persona’s memory and personality
Syncing real-time voice generation with chat for smooth conversations
Handling authenticated + demo users within the same logic paths
Creating an elite, mobile-first UI in a short timeframe

🏆 Accomplishments that we're proud of

✅ Fully functional AI memory companion built in < 1 week
🎙️ Smooth voice AI via ElevenLabs
🔐 Full real-time Supabase backend with Row Level Security
✨ High-quality UI/UX with smooth animations
📱 PWA support: works offline, installable, blazing fast

📚 What we learned

How to integrate multi-modal AI (text, voice, memory) into a cohesive experience
The value of designing with emotion and empathy, not just logic
Supabase’s full capabilities for real-time secure apps
Building a premium-grade PWA on a hackathon timeline

🚀 What's next for Relive

🧠 Persona Studio: visual memory builder using AI to auto-generate personalities
🗣️ Real-time voice-to-voice calling between users and personas
🔐 RevenueCat-powered Pro subscriptions with premium personas
🎥 Full Tavus video agent integration for avatar-based conversations
🕰️ Memory Capsules: send messages to/from past/future versions of yourself
💬 Social sharing of emotional moments via AI-generated clips