Vivid is a beautiful, fast, and intelligent AI-powered Second Brain mobile app. Users capture thoughts, notes, voice memos, images, PDFs, and links — the AI automatically organizes everything, finds connections, and lets users chat naturally with their entire personal knowledge base.

Vivid is submitted to HACKHAZARDS '26 for two tracks:

1. Expo Track — Build Mobile Apps with Expo
2. Sarvam AI Track — Build AI Applications with Sarvam AI

"The first multilingual Second Brain for India — capture thoughts in any Indian language, search and chat with your knowledge base in your mother tongue."

While existing second-brain apps (Notion, Obsidian, Mem.ai, Reflect) are English-only, Vivid breaks the language barrier:

• 🎙️ Voice capture in 10+ Indian languages using Sarvam STT (Speech-to-Text)
• 💬 Chat in your mother tongue — ask questions in Hindi, Tamil, Telugu, and get answers back in the same language
• 🔊 Listen to AI responses in your language with Sarvam Text-to-Speech
• 🌐 Auto-detect and translate between English and Indian languages seamlessly

Vivid leverages the Sarvam AI platform for all multilingual capabilities:

To run Vivid with multilingual features, add your Sarvam API key:

# .env
EXPO_PUBLIC_SARVAM_API_KEY=your_sarvam_api_key_here

Get your API key at: https://dashboard.sarvam.ai

We have successfully implemented all core UI screens and basic functionality. We have now integrated backend services (Supabase) and state management (Zustand) with local database persistence (Expo SQLite + Drizzle ORM) for offline-first capabilities.

Phase Progress:

• Phase 1: Project Setup & Core Navigation
• Phase 2: MVP Foundation Complete
• Phase 3: Backend Integration & AI Enhancement (Local DB Complete, AI Chat Integration Complete, Auth Complete, Syncing Complete)
• Phase 4: Polish, Advanced Features & Deployment (Onboarding Complete, Dark Mode Complete, Settings Complete, Task Management Complete, Graph View Complete, Production Ready)

• React Native + Expo (latest version) with TypeScript
• Expo Router for file-based navigation
• Tab-based navigation (Home, Chat, Library, Settings)
• Privy authentication integration (configured with actual credentials)
• Authentication flow implemented (Email, Google, Apple login)
• Onboarding flow with animations ✅
• Dark mode support with theme provider ✅
• Settings screen with theme toggle, sync, and more ✅

• Knowledge feed showing recent captures
• AI Daily Insight card with generated insights
• Search bar with basic functionality
• Floating Action Button with capture options:
  • Quick Text Note (functional with AI tagging)
  • Voice Memo (functional)
  • Photo/Image Capture (functional)
  • PDF/Document Upload (functional)
  • Link/Article Save (functional)
  • Task Reminders ✅ (mark as task, set reminder date)

• Text Note: Title and content input with AI-powered tagging and categorization
• Voice Memo: Recording with visual timer and audio capture
• Image Capture: Camera and gallery picker with preview
• Document Capture: File picker with type detection
• Link/Article: URL input with preview fetching simulation

• Dedicated chat interface with message history
• Voice input capability (microphone button)
• Real AI integration with Groq (fallback to Gemini)
• RAG (Retrieval-Augmented Generation) using your knowledge base
• Citations and references to specific notes in responses
• Suggested follow-up questions
• Real-time typing indicators
• Loading states and error handling

• Visual network of knowledge connections
• Interactive nodes with emoji indicators
• Zoom in/out controls
• Node selection shows details
• Connections based on shared tags and content similarity

• Organized views by type: All, Ideas, Tasks, Insights, Projects, People, References
• Favorite/bookmark functionality
• Item previews with type indicators and badges
• Stats dashboard showing totals and favorites

• Visual network of knowledge connections
• Interactive nodes with emoji indicators
• Zoom in/out controls
• Node selection shows details
• Connections based on shared tags and content similarity

• Full-screen view of captured items
• Action buttons (Share, Duplicate, Delete)
• Metadata display (type, date, tags)

• Clean, minimalist design with soft dark/light mode readiness
• Smooth animations using Moti
• Beautiful typography and spacing
• Platform-appropriate styling (iOS/Android)
• Glassmorphism effects and subtle depth
• Micro-interactions and feedback
• Responsive layouts

• Framework: React Native + Expo (latest version)
• Authentication: Privy (configured with actual app ID and secret)
• UI: NativeWind (Tailwind) + Moti for animations
• Navigation: Expo Router (file-based)
• State Management: Zustand with persistence (implemented)
• Local Database: Expo SQLite + Drizzle ORM (client configured, schema defined, basic operations implemented)
• Backend: Supabase configured (client and schema ready, integration in progress)
• AI Integration: Groq and Gemini clients configured with fallback logic
• Media: Expo AV (audio), Expo ImagePicker, DocumentPicker
• Icons: Lucide React Native

• Updated app icon and splash screen using the provided Vivid logo
• Both icon and splash screen now use: https://res.cloudinary.com/dv0tt80vn/image/upload/v1777551455/vivid_cynifa.png

vivid/
├── app/
│   ├── _layout.tsx           # Root layout with PrivyProvider
│   ├── (tabs)/               # Tab-based navigation
│   │   ├── _layout.tsx       # Tabs layout
│   │   ├── index.tsx         # Home screen
│   │   ├── chat/             # Chat screen
│   │   │   └── index.tsx
│   │   └── library/          # Library screen
│   │       └── index.tsx
│   ├── capture/              # Capture screens
│   │   ├── text/             # Text note capture
│   │   │   └── index.tsx
│   │   ├── voice/            # Voice memo capture
│   │   │   └── index.tsx
│   │   ├── image/            # Image capture
│   │   │   └── index.tsx
│   │   ├── document/         # Document capture
│   │   │   └── index.tsx
│   │   └── link/             # Link capture
│   │       └── index.tsx
│   ├── knowledge/            # Knowledge item detail
│   │   └── [id].tsx
│   ├── components/           # Shared components
│   │   ├── FloatingActionButton.tsx
│   │   ├── KnowledgeCard.tsx
│   │   ├── AIDailyInsightCard.tsx
│   │   └── SearchBar.tsx
│   └── types/                # TypeScript types
│       └── knowledge.ts
├── components/               # Additional components
├── assets/                   # Images, icons, etc.
├── README.md
├── package.json
├── TODO.md
└── tsconfig.json

1. Install dependencies:

   npm install

2. Configure environment variables:

   • Create .env file with:

     EXPO_PUBLIC_PRIVY_APP_ID=your_privy_app_id
     EXPO_PUBLIC_SUPABASE_URL=your_supabase_url
     EXPO_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key
     EXPO_PUBLIC_GROQ_API_KEY=your_groq_api_key
     EXPO_PUBLIC_SARVAM_API_KEY=your_sarvam_api_key_here
     

3. Start the development server:

   npx expo start

See TODO.md for complete roadmap, but immediate priorities include:

1. Backend Integration

   • Set up Supabase for data persistence
   • Implement offline-first sync with Expo SQLite + Drizzle ORM
   • Replace useState with Zustand/TanStack Query for state management

2. AI Features

   • Integrate Groq API for smart tagging during capture
   • Implement RAG (Retrieval-Augmented Generation) for chat
   • Add structured outputs for consistent AI responses

3. Persistence Layer

   • Implement Drizzle ORM schemas for knowledge items
   • Create CRUD operations for all capture types
   • Add synchronization with Supabase when online

4. Enhanced AI Chat

   • Connect chat to actual AI API (Groq/Gemini/OpenAI)
   • Implement proper context management
   • Add citation/link referencing in responses

5. Advanced Features

   • Task management with reminders (Expo Notifications)
   • Habit/progress tracking
   • Graph/connections view
   • Daily AI-powered notifications
   • Export functionality

See DEMO_SCRIPT.md for the complete 2-3 minute demo video script.

• App Icon: assets/icon.png (Vivid logo with gradient)
• Splash Screen: assets/splash-icon.png
• Screenshots: (To be taken during testing)

1. Hook (0:00-0:20): "Your notes, but smarter"
2. Capture (0:20-0:50): Text, voice, photo, document, link
3. AI Chat (0:50-1:30): THE KILLER FEATURE - Ask Vivid anything
4. Graph View (1:30-1:50): Visual connections
5. Advanced (1:50-2:10): Task reminders, dark mode
6. Closing (2:10-3:00): Call to action

Record using: Actual phone + screen recording for authenticity!

This is a hackathon project. Feel free to fork and experiment!

MIT